WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Post Office Horizon IT Inquiry Reference: WITN1039R9-1/AP/20231212
Witness Name: Andrew Paul Parsons
Statement No: WITN10390200
Exhibits: See Annex 1
Dated: 17" April 2024
POST OFFICE HORIZON IT INQUIRY
FIRST WITNESS STATEMENT OF ANDREW PARSONS
Il, Andrew Parsons, of Oceana House, 39-49 Commercial Road, Southampton
$015 1GA, will say as follows:
A. INTRODUCTION
1. 1am a Solicitor and Partner in the Commercial Litigation Practice of Womble
Bond Dickinson (UK) LLP (“WBD?’ or the “firm”). I make this statement to assist
the Post Office Horizon IT Inquiry (the ‘Inquiry”) with the matters set out in its
Rule 9 Request dated 12 December 2023 (the “Request’), relating to Phase 5
of the Inquiry’s work.
2. As the Inquiry will be aware, the subject matter of the Request relates to work
which I undertook on behalf of Post Office Limited (POL) as an external
commercial litigation solicitor over a period of approximately seven years
beginning (now) more than 10 years ago. As I set out further below, that work (in
very broad overview) comprised: assisting POL in preparing its responses to
Page 1 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
“Spot Reviews” conducted by Second Sight Support Services Limited (“Second
Sight’) in Spring to Summer 2013; later in 2013, advising POL on the
establishment of the Initial Complaint Review and Mediation Scheme (the
“Mediation Scheme’); representing POL during the period of operation of the
Mediation Scheme through to the closure of the Mediation Scheme Working
Group (the “Working Group”) in 2015; and, from early 2016 onwards,
conducting POL’s defence to the claims brought against it by a group of sub-
postmasters and sub-postmistresses (“SPMs”) in the litigation known as Alan
Bates & Others v Post Office Limited (the “group litigation”), which ultimately
settled in December 2019.
3. For convenience, I refer to this work compendiously as the “Horizon-related
matters”, however this is not intended to imply that these were POL’s only
matters in which the Horizon IT system was or may have been relevant; or that
these were the only such matters in which I or WBD were involved; or that the
above brief summary is an exhaustive description of all the work which I
undertook for POL between 2013 and 2019. Nor is the use of this shorthand
intended to suggest that the work I carried out for POL in this period formed part
of a single overarching or continuous brief. Although I had an ongoing
relationship with POL throughout this period, it would be more accurate to
characterise the Horizon-related matters as an evolving series of instructions,
the scope and nature of which varied. I address this further below.
4. The range, scale and duration of the Horizon-related matters, and the passage
of time since much of this work was done, means that my recollection of the
specifics of many of the events and documents referred to in the Request is very
limited. In some instances I have a general, high-level recollection of the matters
Page 2 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
referred to, but on the whole I have been heavily reliant on reviewing
contemporaneous documents in order to answer the Inquiry’s questions — and,
indeed, I have often found that I can do no more than review the relevant
documents and set out what is or appears to be recorded therein. In each case
where I rely on documents in order to answer a question, I have considered them
against what I am able to remember in order to ensure that the meaning I have
ascribed to them represents my best and honest evidence. Equally, where I am
able to independently recall a particular event or document, I draw this to the
Inquiry’s attention below. Accordingly, I have broadly adopted the following
protocol when setting out my evidence:
4.1. Where I have a memory that stands out independently of the documents,
I say, “/ remember’;
4.2. Where I have a recollection that has been aided or refreshed by reference
to the documents, I say “/ recall’ or “my recollection is”;
4.3. Where I say something like “/ believe”, I generally mean that I am drawing
an inference from the documents and that I believe that inference to be
accurate based on all that I can, to the best of my ability, remember;
4.4. Otherwise, the matters set out in this statement are based solely on the
documentation available to me (unless they are drawn from some other
source, in which case I say so).
5. I also wish to highlight that my firm has a very large amount of documentation on
file concerning the Horizon-related matters. Indeed, I understand that my firm
holds over 900,000 documents comprising the contents of POL files that we
believe may have some link to Horizon-related matters. In addition, my email
Page 3 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
account holds a large amount of relevant material. For the avoidance of doubt
these figures do not describe unique documents — there are a significant number
of duplicates within them, including between my email account and our files —
and my email account will contain documents not relevant to POL. It was
impossible for me to review them all in order to prepare this statement, so I have
focused on the questions and documents which have been put to me in the
Request. Targeted searches for further documents directly relevant to the
Inquiry’s questions have been conducted by myself and my legal advisors.’
Where I have identified documents pertinent to the Inquiry’s questions that are
not referred to in the Request I have endeavoured to draw them to the Inquiry’s
attention, however I wish to stress that there will be a significant number of other
relevant documents which it has not been practicable for me to review or refer to
(including, inevitably, documents which I had sight of at the time of the events
under consideration but have since forgotten). Through a combination of my
limited memory, the breadth of the events involved, the number of documents
and the limited time available for the preparation of this statement, there is a real
possibility that material points may have been missed, but I have done my best
to present the Inquiry with a complete and accurate picture.
6. The remainder of this statement is structured as follows:
6.1. Section B (§§9-14) addresses various preliminary matters, including the
process by which this statement has been prepared and my approach to
privileged documents.
1 Below, where I refer to a search ‘I have carried out’ I mean a search carried out either by me
personally or by one of my legal advisers acting on my behalf.
Page 4 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
6.2. Section C (§§15-30) briefly sets out my professional background, gives
an overview of my work for POL prior to the Horizon-related matters, and
sets out the proportion of my time that was spent on these matters whilst
they were underway. Broadly speaking this section addresses Q1 and Q3
to Q5 of the Request.
6.3. Thereafter I attempt to deal with the matters raised in the Request in
chronological order, whilst also adhering, so far as practicable, to the
structure of the topics set out therein. As the Inquiry will appreciate, this is
not straightforward as a number of topics concern similar or overlapping
periods, and because some of the questions asked are thematic rather
than chronological.
6.4. Section D (§§31-50) concerns the assistance I provided to POL during the
period when POL was preparing its responses to the Spot Reviews and
Interim Report produced by Second Sight in Spring to Summer 2013. This
section addresses Q2 and Q10 to Q15 of the Request.
6.5. Section E (§§51-77) concerns POL’s response to Second Sight's interim
report dated 8 July 2013 (the “Second Sight Interim Report” or the
“Interim Report”), and the advice and assistance I provided to POL in
establishing the Mediation Scheme following publication of the report. The
scheme opened to new applications on 27 August 2013 and closed on 18
November 2013. This section responds to Q16 to Q19 of the Request.
6.6. The period covered by Section F (§§78-137) overlaps with the periods
covered by Sections D and E. In broad terms, this section concerns
various matters relevant to the work which POL was carrying out at that
Page 5 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
time to review historic prosecutions of former SPMs which had relied at
least in part on Horizon data. This section addresses:
(i) First, and in overview, the nature of my role as a civil lawyer acting for
POL and the extent to which I was sighted on these matters (§§78-81);
(ii) Second, my receipt in early July 2013 of a report in respect of the
Lepton SPSO by Helen Rose, a fraud analyst at POL (the “Helen Rose
Report’), and the review of criminal cases which the report triggered
(§§82-89; Q20 of the Request);
(iii) Third, POL’s response to correspondence sent by the Criminal Cases
Review Commission (“CCRC”) in July 2013 following the publication of
Second Sight’s Interim Report, and its ensuing decision to appoint
Brian Altman QC to supervise the review of criminal convictions and
advise on strategy in relation to the CCRC (§§90-97; Q29 and Q31 of
the Request);
(iv) Fourth, my participation in what I refer to as the “Horizon Regular
Calls” during the second half of 2013 and at the beginning of 2014 (the
last such call I attended being, to the best of my knowledge, the one
on 19 February 2014) (§§98-122; Q21 to Q28 of the Request);
(v) Fifth, the conference with Brian Altman QC which I attended on 9
September 20 (§§123-136; Q30 of the Request).?
6.7. Section G (§§138-259) deals with the period during which the Mediation
Scheme had closed to new applications and was underway (i.e. November
2 For convenience, I also deal with Q33 of the Request at this juncture.
Page 6 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
2013 until mid-2015). Mediations in fact continued after mid-2015 but my
work on the scheme was substantially complete by then.
(i) First, at §§139-155, I give an overview of my (and more broadly
WBD’s) role in relation to the Mediation Scheme during its period of
operation, including: my role in attending meetings of the Mediation
Scheme Working Group on behalf of POL (Q34 to Q35 of the Request);
WBD's role in preparing the “Horizon Factfile” document (Q36 of the
Request); and the nature and extent of WBD’s role in relation to POL’s
investigations into individual cases within the Mediation Scheme (Q38
of the Request).
(ii) Second, I address various questions about the process POL followed
when investigating applicants’ complaints, namely: Q37, Q39 to Q40,
and Q42.1 (§§156-162).
(iii) Third, I answer various questions raised by the Inquiry in Q41 and
Q42.2 to Q44 of the Request about my developing views on the work
carried out by Second Sight during the Mediation Scheme (§§163-
175).
(iv) Fourth, I address various questions the Inquiry has asked about the
provision of certain documents or other kinds of information during the
Mediation Scheme — including the Helen Rose Report (cf. Q32 and
Q49; §§176-184); reports of investigations into suspected criminality
by POL investigators (“Officer’s Reports”, cf. Q46; §§185-201); and
Page 7 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
information about a form of remote access in Horizon Online known as
the “Balancing Transaction” functionality (cf. Q45; §§202-232).5
(v) Fifth, at §§233-250, I set out my (and WBD’s) role in advising POL on
the merits of applicants’ cases and whether or not to take a case to
mediation. This answers Q38, as well as Q47 to Q48.
(vi) Sixth, at §§251-259, I set out my response to Q35.6 and Q50 to Q52
of the Request, concerning POL’s approach to deciding whether or not
to take a case to mediation, and the ultimate decision to mediate cases
and close down the Working Group.
6.8. Against this background Section H (§§260-282) summarises the extent of
my knowledge, during the periods covered by the preceding sections, of:
(i) bugs, errors and defects in Old Horizon and Horizon Online; and (ii)
remote access in Old Horizon and Horizon Online. In order words, this
section distils my response to Q6 to Q9 of the Request relative to the
period up until my work on the Mediation Scheme largely ended in mid-
2015 and prior to the instigation of the group litigation in early 2016.
6.9. Section I deals with Q53 to Q54 of the Request (§§283-289).
6.10. Section J (§§290-308) sets out the extent of my awareness of and
involvement in the “Swift Review” which was carried out by Jonathan
Swift QC (as he then was) and Christopher Knight between late 2015 and
February 2016 (Q55 to Q57 of the Request).
3 This subsection also partly addresses Q88 of the Request, insofar as it describes my awareness
of and involvement in “Project Zebra” at the time it was carried out in c.2014.
Page 8 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
6.11. Section K (§§309-393) sets out, in overview, what my role was in relation
to the group litigation; how the group litigation was managed; from whom
I received instructions; and my relationships with key personnel at POL,
the Counsel team, and those representing the Claimants. It also
summarises the advice I gave POL in relation to its general litigation
strategy and tactics. These are broadly speaking the matters raised at Q58
to Q63, Q67 and Q70 of the Request.
6.12. Section L (§§394-451) deals with various questions raised by the Inquiry
in relation to the early work which I/WBD carried out on the group litigation
(mostly prior to service of POL’s Generic Defence in July 2017). These
questions are centred around the related themes of preservation of
documents, early requests for disclosure, and other forms of information-
sharing, as follows:
(i) Advice given advice in relation to the preservation of documents (Q64).
(ii) POL’s response to a Data Subject Access Request (“DSAR’) in April
2016 by one of the Claimants (Q65).
(iii) POL’s response to the Claimants’ request (made in their Letter of Claim
dated 28 April 2016, or “LOC”) for early disclosure of its internal
investigation guidelines (Q68).
(iv) Information given to Leading Counsel, Tony Robinson QC, in the
course of instructing him in May 2016 (Q69).
(v) Advice given by Tony Robinson QC in conference on 9 June 2016, in
particular in connection with preserving privilege in the implementation
of the Swift Review (Q72).
Page 9 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(vi) Advice given to POL about sharing information on the group litigation
with UK Government Investments (“UKGI”) and the Department for
Business, Energy and Industrial Strategy (“BEIS”): Q71. This largely
arose later in the litigation but is dealt with in Section L for convenience.
6.13. The matters covered in Section M (§§452-520) overlap in time with
Section L, but relate to early investigative and preparatory work rather than
work on documents and information-sharing. Section M therefore covers:
(i) First, the work undertaken to prepare POL’s Letter of Response dated
28 July 2016 (the “LOR”) to the LOC, including early investigative work
by Deloitte into the issue of remote access as part of “Project
Bramble”: Q73 to Q74 (§§454-478).
(ii) Second, the work undertaken to prepare POL’s Generic Defence
served around a year later on 18 July 2017, including further
investigative work by Deloitte into remote access as part of “Project
Bramble”: Q75 (§§479-509).
(iii) Third, the advice I gave POL upon receipt of Deloitte’s draft report for
Project Bramble in September 2017: Q89 (§§510-520).
6.14. Section N (§§521-694) concerns disclosure. It covers a range of topics
related to the management of the disclosure process on behalf of POL, in
particular the disclosure orders made at the CMCs from October 2017 to
June 2018, and the approach to disclosure thereafter. In summary, the
topics addressed are:
(i) General advice on disclosure (Q58.4; §§523-551);
Page 10 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(ii) Disclosure of the Known Error Log (“KEL”) database (Q76 to Q82;
§§552-598);
(iii) Disclosure of the ‘Peak’ database (Q81.2 and 83 to Q87; §§599-647);
(iv) Disclosure of the reports generated by Project Zebra (Q88.3; §§648-
650);
(v) The approach taken to redacting evidence, including that deployed in
the Common Issues Trial and Horizon Issues Trial (Q90.1, Q91 and
Q95.1; §§651-671); and
(vi)Events surrounding the obtaining of certain audit documents held by
Royal Mail for the purposes of the Horizon Issues Trial (Q99; §§672-
694).
6.15. Section O (§§695-768) answers the Inquiry’s questions about POL’s
preparation for the Common Issues Trial which took place over 15 non-
consecutive days in November and December 2018. These questions are
broadly Q90 and Q92 to Q94 of the Request (Q90.1 and Q91 being dealt
with in the preceding section on disclosure).
6.16. Section P (§§769-912) answers Q95 to Q102 of the Request (save for
Q95.1 and Q99, which is dealt with in the previous section on disclosure).
These questions concern POL's preparation for the Horizon Issues Trial,
which took place over 21 non-consecutive days between 11 March and 2
July 2019, including: the preparation of the witness statements of certain
employees of Fujitsu; my/WBD’s involvement in the preparation of POL’s
expert evidence; and the basis for POL’s continued belief, going into to
that trial, that the Horizon system was ‘robust’ (and to this extent, this
Page 11 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
section summarises my response to Q6 to Q9 of the Request as regards
the period covered by the group litigation).
6.17. Section Q (§§913-989) deals with POL’s response to Mr Justice Fraser's
judgment [2019] EWHC 606 (QB) (the “Common Issues Judgment’),
handed down on 15 March 2019 shortly after the start of the Horizon
Issues Trial, including: (i) POL’s decision to seek leave to appeal that
judgment; and (ii) the application made by POL on 21 March 2019, seeking
the recusal of Mr Justice Fraser as the Managing Judge in the group
litigation and the adjournment of the HIT which was then underway (the
“Recusal Application’). In broad terms, therefore, this section addresses
Q103 to Q118 of the Request.
6.18. Section R (§§990-1006) summarises events after the conclusion of the
Horizon Issues Trial, and also briefly deals with the matters raised at Q119
to Q120 of the Request.
6.19. Finally, in Section S (§§1007-1016) I make some observations by way of
overview and conclusion. I do so in an effort to assist the Inquiry and to
answer the questions posed at Q121 to Q122 of the Request.
7. Before I turn to the matters outlined above, I should say something about my
approach to the work I did for POL. As a solicitor acting for a client involved in
civil disputes, it was my role to advance POL's interests to the best of my ability
andin line with my professional ethics. That inherently meant advancing
positions that were sometimes adverse to the interests of SPMs. I was and am
acutely aware of the consequences of doing that, but as a solicitor that was my
duty. Over the years of our acting for POL, my firm and I did make a few mistakes
Page 12 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
as I set out in this statement. They were genuine errors and at all times I believe
that my firm and I acted appropriately. I apologised for them at the time and
apologise for them now. A great deal of information about the Horizon system
has come to light over the years and, like everyone, I now know that there have
been miscarriages of justice. Although, I was not myself involved in advising
POL on prosecutions, I am concerned to help the Inquiry so far
as I can in its important work to understand how this happened.
8. In this statement, I shall provide an account and answer the Inquiry’s questions
to the best of my ability. Of course, I can only give answers about matters that
were and are within my knowledge (as explained at §4 above). In that regard, I
should note that there were many other lawyers in my firm who played a role at
particular stages (which at its peak exceeded 20 qualified lawyers, plus
paralegals) and so I was not involved first-hand in all events. I give some details
below where their work is relevant. I should also observe that POL had in-house
lawyers and used other firms, so WBD and I were not always aware of work being
done and advice being received from others.
B. PRELIMINARY MATTERS
9. I address some preliminary matters before turning to the substance of the
questions set out in the Request in the sections that follow.
(i) Process by which this statement was prepared:
10. In preparing this statement I have been assisted by my legal team, which is made
up of lawyers and paralegals within WBD as well as external Counsel. As the
Page 13 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Inquiry will appreciate, the preparation of this statement has involved, of
necessity, reviewing a significant amount of material. In large part, I was reliant
upon input from my legal team to access documents and to carry out searches
for documents. The evidence contained within this statement, as a result of that
review process, is mine and mine alone.
(ii) People and documents referred to in this statement:
11. Documents referred to in this statement which were provided to me by the Inquiry
along with the Request are listed in Annex 1 in the order in which they are
referenced.
12. Documents referred to which did not form part of the Request, but which I have
identified from my firm’s systems in the course of preparing this statement, are
also listed in Annex 1 and have been provided to the Inquiry. These documents
are also listed in the order in which they are referenced.
13. A large number of people are referred to in this statement, including (but not
limited to) employees of POL, Fujitsu and WBD. Where I do so, I refer to them
by the names, titles and job titles they held at the material time (for example, “Mr
Justice Fraser” instead of “Lord Justice Fraser”, and “QC” instead of “KC”).
(iii) Privilege:
14. Iam aware that POL has waived legal professional privilege in respect of certain
matters and documents (“POL’s waiver’). As a result, a number of documents
which would otherwise be privileged have been produced to me along with the
Request (including some documents from my firm's systems which were
Page 14 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
previously provided to POL for onward disclosure to the Inquiry). I am aware of
the terms of POL’s waiver and have done my best to apply it in answering the
Inquiry’s questions and providing documents from my firm’s records along with
this statement. As will be seen below, there are only a limited number of areas
where the ambit of POL’s waiver is relevant to my responses, principally Section
C and my answer to Q120.
C. BACKGROUND INFORMATION (Q1, Q3 to Q5)
(i) Professional background
15. I studied law at the University of Warwick, graduating in 2005. After graduating I
took the Legal Practice Course (“LPC”) and joined the firm (which was then Bond
Pearce) as a trainee in September 2006. Bond Pearce subsequently merged with
Dickinson Dees in May 2013 to become Bond Dickinson, and later (with effect
from October 2017) combined with a US-based law firm, Womble Carlyle
Sandridge & Rice LLP, to form Womble Bond Dickinson. For convenience,
throughout this statement I refer to WBD and its predecessors compendiously as
“WBD’ or the “firm”.
16. I completed my training contract and qualified in September 2008, joining the
firm’s Commercial Litigation Practice as a solicitor in its Southampton office. As
a junior solicitor I undertook a range of general commercial litigation work but
within a few years I became increasingly focused on IT, digital and technology
issues, including dealing with data protection and privacy matters, disputes
arising out of the provision of IT services, and disputes arising out of hacking and
Page 15 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
other security breaches. I was promoted to Senior Associate* by 2012 before
becoming a Partner in May 2016.
17. I have specialised in commercial litigation for the whole of my career and have
never specialised in criminal law. Whilst my work has occasionally had some
crossover with criminal law aspects (for example, where an alleged misuse of
personal data may amount to a criminal offence, in addition to giving rise to civil
liability), I do not regard myself as a criminal lawyer and I do not have any
experience of bringing or defending criminal prosecutions. Nor, prior to 2013
when I was first instructed by POL in relation to the matters under consideration
by the Inquiry, did I have any knowledge or experience of the rules governing
disclosure (including post-conviction disclosure) in criminal proceedings — save
to the extent that this may have been covered in the compulsory criminal law
module that I studied during the LPC. As I set out further below, over time I gained
some exposure to these topics through my work for POL and hearing advice from
POL's criminal lawyers, but I still do not regard criminal disclosure as being within
my field of expertise. Those instructing me at POL were aware of this at all times.
18. My work has always mainly consisted of commercial dispute resolution for private
clients, such as large corporate clients. Other than my work for POL I have not
undertaken a significant amount of work for public authorities or state-owned
companies; my experience has predominantly been acting for private entities in
commercial disputes. I may have advised other public bodies or state-owned
companies from time to time (1 do not recall any specific examples but my firm
does generally act for such organisations), including prior to 2013 when I was
4 At some point Senior Associates in the firm were renamed Managing Associates.
Page 16 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
first instructed in relation to the matters referred to in the Request, but this would
have been a very small part of my work and did not concern anything substantial.
19. I am asked whether I had “experience of group litigation concerning a major IT
project’ when I was first instructed by POL in relation to the matters under
consideration by the Inquiry (Request, Q3.3). I did not have any prior experience
of group actions, either in respect of major IT projects or otherwise. However, to
my knowledge Bates was the first group action focused on a major IT system in
English legal history. I therefore had no prior experience in this specific type of
case either at the time of my initial instruction by POL, or indeed subsequently
when the group litigation was initiated, because it was the first case of its kind.
(ii) POL as a client prior to the Horizon-related matters
20. I am asked to what extent, if at all, POL was seen within WBD as “either (a) an
important client or (b) a good source of future work” (Request, Q4). I interpret
this as referring to WBD’s perception of POL as a client prior to the Horizon-
related matters.
21. As I set out further below in Section D, I first became instructed in relation to
Horizon-related matters in around April 2013, when I was asked to assist POL in
preparing its responses to Spot Reviews then being conducted by Second Sight.
22. Prior to this I recall that I had acted on certain small matters for POL (which here
also refers to Royal Mail Group or “RMG” prior to their separation on 1 April
2012). When I joined WBD in 2006, POL was already a client of the firm and
when I qualified in 2008, the firm was engaged in bringing claims against SPMs
for shortfalls in branch accounts. Claims were dealt with by a team of paralegals
Page 17 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
in the firm's Plymouth office unless they were either contested or of a higher
value, in which case they were dealt with by lawyers in the commercial litigation
team, who were generally below partner level due to their (still) comparatively
low value.
23. Between 2008 and approximately 2011, I was engaged by POL to act in a handful
of these shortfall claims. My recollection is that my instructions on these matters
largely came from Mandy Talbot (a solicitor in RMG’s legal services team). I do
not recall these matters involving challenges to the integrity of the Horizon
software by the SPMs concerned, and to the best of my recollection none of them
resulted in proceedings being issued (but I have not reviewed those case files
for the purpose of preparing this statement).
24. I was not involved in POL’s claim against Lee Castleton which was tried in
December 2006 (at which point I would have been a couple of months into my
training contract). I read the High Court’s judgment in the course of my
engagement on shortfall claims because the Court’s decision that SPMs could
be pursued in debt on the basis that they had submitted a “settled account” was
relevant to these claims. I was aware it was a case handled by WBD.
25. I recall that I also acted, in 2011 and 2012, in a couple of matters which involved
working with Susan Crichton and Hugh Flemington, General Counsel and Deputy
General Counsel for POL, respectively. I do not believe POL has waived privilege
in relation to these matters so I say no more about them here.
26. Imay have done some other work for POL prior to 2013, but these are the matters
which I can recall. Over the early years of my practice, I worked for POL on and
off and they were not a key client of mine. I was aware that that they were a key
Page 18 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
client of the firm, but I do not recall ever being involved in client relationship
management nor that there was anyone from POL who would call me directly for
advice or seek to instruct me directly.
27. At the start of 2012, POL and RMG combined were the largest client of Bond
Pearce (as it then was). On 1 April 2012, POL and RMG separated, and
subsequently the firm principally worked for POL; our work for RMG reduced
substantially after that point. Additionally, and as set out above, in 2013 Bond
Pearce and Dickinson Dees merged. Thus, through a combination of POL and
RMG being separate entities and the firm growing in size, POL was a relatively
smaller, but still major, client of the firm from 2012-2013. It continued to be seen
within the firm — including by me — as an important client, in the same way that
all other clients of the firm are considered important. Similarly, POL was
considered to be a good source of future work in the same way that the firm aims
to retain all of its clients for future business.
(iii) Proportion of my work devoted to the Horizon-related matters
28. Q§5 of the Request asks “what proportion of [my] work related to acting for POL
in matters concerning the Horizon IT System from [my] initial instruction to ...
ceasing to act on the matter’. There are three initial points to make in relation to
this. First, as I have explained, I identify the beginning point of my work on the
Horizon-related matters as being around April 2013, when I was engaged to
assist POL in preparing its responses to Spot Reviews. Second, as mentioned
at §3 above, the nature and scope of my engagement for POL varied thereafter,
and it is not accurate to characterise the Horizon-related matters as a single or
continuous “matter”. Third, it is difficult to identify an end point with precision but
Page 19 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
29.
for present purposes I adopt the end of 2019 shortly after the group litigation
settled (and in any event I note that POL has not waived privilege in relation to
matters after 26 February 2020).
Subject to those caveats, for the purpose of answering Q5 I consider it to be
useful to break down my involvement in the Horizon-related matters into three
broad phases:
29.1. Phase 1 (April to July 2013): during this period I, along with other
colleagues at WBD, assisted in the drafting of POL’s formal responses to
Spot Reviews. I had day-to-day conduct of this matter, supervising the
work of more junior solicitors as well as assisting with some drafting
myself. Overall, and in broad terms, I estimate that this work represented
about 10 per cent of my total workload.
29.2. Phase 2 (July 2013 to early 2016): during this period, I, along with other
colleagues at WBD: assisted POL in preparing for and responding to
Second Sight's Interim Report; advised POL on the establishment and
running of the Mediation Scheme; assisted POL with drafting investigation
reports during the Mediation Scheme; and represented POL at mediations.
Again, I was the solicitor at WBD with day-to-day conduct of this
instruction. I would estimate that work on Horizon-related matters during
this represented around 50 per cent of my workload on average. From
around mid 2015 there was a substantial reduction in work as the
Mediation Scheme was running down and work on the group litigation did
not begin in earnest until early 2016; during this time I estimate that my
work for POL on Horizon-related matters ultimately reduced to around 10
per cent of my total workload.
Page 20 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
29.3. Phase 3 (early 2016 until the end of 2019): during this period, the group
litigation was in full swing. I became a Partner in May 2016, and was the
lead Partner with conduct of the litigation on behalf of POL. Typically
around 70 per cent of my time related to this instruction but there were
discrete periods where 100 per cent of my time was devoted to it.
30. It should be noted that the foregoing estimates are based on my general
impression of the relative amount of time I spent working on the Horizon-related
matters, and not (for example) on a detailed examination of historic timesheets.
D. SECOND SIGHT SPOT REVIEWS (Q2, Q10 to Q15)
(i) Instructions to assist with Spot Reviews
31. As mentioned above, I first became instructed in the Horizon-related matters in
around April 2013 when I was a Senior Associate. As I later learned through
discussions with people at POL, in around June 2012, POL had appointed
Second Sight to conduct an independent investigation into whether there were
systemic problems with the Horizon system (including training and support
processes) pursuant to terms of reference which were (I believe) agreed between
POL, the Justice for Sub-postmasters Alliance (“JFSA”) and Second Sight. It
follows from the timing of my instruction in April 2013 that I had no role in advising
POL as to how to respond to the May 2009 Computer Weekly article or
allegations by MPs. Nor did I advise on the decision to appoint Second Sight,
their terms of reference, or the process by which they would carry out their
investigation (including the provisions of the terms of reference determining the
extent of the documentation and information which they would be permitted to
Page 21 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
access). I cannot comment on whether other individuals in my firm may have
done so, but I am not aware of anyone having played any significant role in this
regard.
32. The part of Second Sight’s investigation in which I was initially involved was
conducted by way of what it called “Spot Reviews”. Although Second Sight’s
investigation was not a mediation or arbitration process and was not intended to
resolve specific disputes, individual SPMs could raise concerns about the
Horizon system with JFSA (or by approaching Second Sight directly). Second
Sight would then conduct a “fast track” review of the information provided by
SPMs to identify what the key issues raised were. The key issues that Second
Sight felt merited further investigation were then separated into individual “Spot
Reviews”, meaning that where an individual SPM submitted information which
raised multiple key issues, this could give rise to multiple Spot Reviews.° The
idea was that Second Sight's investigations into these Spot Reviews would
inform its ultimate conclusions as to whether there were systemic problems with
Horizon, training and/or support. I recall there were around 50 SPMs who came
forward with concerns which fell within the scope of Second Sight’s review. Some
of them approached the JFSA and some were referred to the investigation by
their local MPs.
33. Second Sight would send the Spot Reviews to POL to review and respond to, to
assist it with its investigation. By the time I was instructed, POL had been asked
to comment on (I believe) 10 Spot Reviews. This is what prompted POL's
engagement of WBD to assist with the process of preparing its responses.
5 This is set out in Second Sight’s Interim Report, POL00099063.
Page 22 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Specifically, around the beginning of April 2013 POL had invited Gavin Matthews
(then a Partner at WBD) to attend a meeting at POL’s offices in Old Street to
discuss the Spot Review process.® As far as I can recall, I ended up attending on
my own as Gavin could not go (though whether the original intention was for us
both to attend, or whether I stood in for Gavin, I cannot now remember). My
recollection (though I may have this confused with another meeting on this topic)
is that there were approximately ten attendees including Susan Crichton, Rodric
Williams (a litigation lawyer in POL’s legal team), Stephen Baker (POL's Head of
Business Change), Angela Van Den Bogerd (Head of Partnerships at POL),
Gareth Jenkins (Distinguished Engineer, Business Applications Architect at
Fujitsu and the principal architect of the Horizon software). I cannot recall
whether there were other representatives of Fujitsu present.
34. I recall the outcome of the meeting was that POL and Fujitsu agreed that POL
would provide the Spot Reviews submitted by Second Sight to Fujitsu. Where
the Spot Review raised technical issues, Fujitsu would prepare a written note in
response. Where the Spot Review raised issues which were less technical,
Fujitsu would prepare a less detailed note or the response would be prepared by
POL alone.
35. WBD was instructed to assist with this process. Initially my role was to assist in
writing up POL's formal response to each Spot Review to be provided to Second
Sight. After Fujitsu and/or POL had produced their note (or notes) in response to
the relevant Spot Review, POL then provided this material to me, and my job was
to turn it into a formal response which POL could then circulate to Second Sight.’
° As is reflected in this email dated 8 April 2013: WBON0000726.
7 See for example POL00098035.
Page 23 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Further information and comments on draft formal responses would sometimes
be provided via email or in calls / meetings attended by representatives of POL
and (at least some of the time) Fujitsu. I do not recall having direct, unsupervised
(i.e. unsupervised by POL) access to Gareth Jenkins or other Fujitsu employees
in the course of this work. With reference to the Inquiry's Q13 and Q14.3 to
Q14.4, it was not part of my role to provide POL with strategic advice in relation
to the Spot Review process, nor do I recall having any particular views on the
investigative process or the quality of the work that Second Sight were carrying
out at this stage. I did not have any direct contact with Second Sight at this point
and had little visibility of communications between them and POL (essentially,
this was limited to seeing emails between the two relating to particular Spot
Reviews when these happened to be forwarded to me). I did not advise on what
disclosures should be made (or not made) to Second Sight at this stage.
36. I do recall having the impression that POL were taking their responsibility to
consider and respond to the Spot Reviews seriously. It appeared to me that care
was being taken to investigate the matters raised by each Spot Review internally
and with Fujitsu, and on their face the Spot Review responses appeared to be
thorough and considered.® Similarly, the Fujitsu papers with which I was provided
so as to draft the formal responses seemed to me to be detailed and carefully
put together, and it appeared from Gareth Jenkins’ comments and answers to
queries that he was interested in ensuring that the Spot Review responses were
accurate, thorough, and supported by technical analyses. That said, it was not
8 A view which appeared to me to have been shared by Second Sight at the time. Its feedback on
the first batch of Spot Review responses (forwarded to me by Simon Baker on 17 May 2013)
included comments that they were “well-written” and “thorough” (POL00098294); Second Sight
also remarked on the “thoroughness” of POL’s investigations in relation to the Spot Reviews
submitted to it in its Interim Report at e.g. paragraph 7.3: POL00099063.
Page 24 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
part of my role to probe the information provided by POL and Fujitsu which fed
into the responses. Further, the responses themselves were complex and I did
not have a firm grasp of their substance (nor was it part of my instructions to do
so). To the best of my knowledge, I did not receive a detailed written briefing from
POL on the operation of Horizon or the background to the Second Sight
investigation prior to (or following) the meeting on 10 April 2013. I remember
having to pick up what was being discussed at the meeting as it went along, and
I recall the basic differences between Old Horizon and Horizon Online being
explained to me at some point, as well as the role and significance of the Core
Audit Process. But by and large I acquired knowledge about the Horizon system
and the nature of the issues being raised by SPMs progressively over time.
37. In short, my initial instructions in relation to the Spot Review process were a
typical, associate-level, drafting job of a routine ‘hold the pen’ nature. I was
supervised in this work by Gavin and my instructions predominantly came from
Rodric Williams (although I was also in contact with others at POL, for example
Simon Baker and Steve Allchorn, who acted as conduits for the information from
POL and Fujitsu which fed into the Spot Review responses; as well as others
who provided comments on the draft responses). In due course I had some
limited assistance from one or two more junior solicitors in the firm, whose work
I then supervised.°
(ii) Additional work in relation to Spot Review 5
® WBON0000736.
Page 25 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
38. Work over and above that outlined above was occasionally required during this
phase in order for POL to provide a fulsome response to the Spot Reviews. In
particular, in relation to Spot Review 5 (which concerned an allegation that a
former SPM, Michael Rudkin, had observed a POL employee based in Fujitsu’s
Bracknell office demonstrate an ability to pass transactions directly into Horizon),
further investigation was undertaken by POL following the submission of its initial
response to Second Sight. This culminated in (i) a reworked response being
provided and (ii) the taking of a witness statement from the employee who was
the subject of the allegation.
39. POL’s initial response to Spot Review 5 had been handled by POL in-house and
sent to Second Sight on 6 June 2013.'° The response noted the allegation made
by Michael Rudkin and went on to answer the questions raised by Second Sight
in the Spot Review as follows:
“Question 1: What capabilities did the POL Bracknell team have? (As far
as TC or Rem Out type transactions or Journal adjustments are
concerned).
Response: The POL Bracknell Team have no access to the live system
so can conduct none of these transactions.
Question 2: What were the PHYSICAL or LOGICAL controls over their
use of the systems available to them?
Response: There is no Physical or Logical connection to the live system
from the areas in Bracknell being discussed/ investigated. Detailed
documentation has been supplied of the testing processes and
procedures recently audited and the design documents to support this
position.
1° WBON0000343.
Page 26 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Question 3: What audit trail is available to show the extent that they
posted TC or Rem Out type transactions, or Journal adjustments?
Response: When any transactions are posted to the database they are
contained in the audit trail. As both the original Horizon and replacement
HNGx test systems were available to the test teams in that period the
test area and the test data is often refreshed and changed it would not
be possible to identify any transactions from this period in the test
system. Specifically we do not keep audits of test systems, only the Live
system. As stated in response to question 1, the teams in the area of
Bracknell concerned would have no access to the live system.
Question 4: Can we reply on the COMPLETENESS of the audit trail? i.e.
does it record all transactions or just transactions meeting certain
criteria? Is it protected from user manipulation?
Response: The detailed answer to this is included in two papers Horizon
Data Integrity and Horizon Online Data Integrity for Post Office Ltd which
have been presented as evidence in a number of previous court cases.
Question 5: What USER ID was used if TC type transactions or journal
adjustments were posted?
Response: On the old Horizon System (which was Live in 2008) and [sic]
Data introduced to the system in the Data Centre would not be marked
with any user ID.
Question 6: Could the POL Bracknell team log on with either super user
or SMPR credentials?
Response: Not in the live system, see test user policy. See the Horizon
Data Integrity and Horizon Online Data Integrity documents for details.
Question 7: How would TC, Rem Out or Journal Adjustment type
transactions executed by the POL Bracknell team be seen by SPMR of
Branches affected by those actions?
Page 27 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Response: For the PO Bracknell team the SPMR would never see any
changes as they are in the test not live systems.”""
40. Upon receipt of this response Second Sight asked follow-up questions,'? which
were then put by POL to Fujitsu.'° Fujitsu provided information in response which
was then forwarded to me together with information sourced internally from within
POL.
41. For example, Steve Allchorn (who was my main point of contact in connection
with POL’s further investigations into Spot Review 5) sent me the following
unattributed information (in bold) on 19 June 2013, which I infer had been
obtained from Fujitsu:
“One of the further challenges asked was despite there not being a
capability to interact with the live Horizon from the test area in the
basement at Bracknell, could it take place if someone had a criminal
intent to hack the system from the basement. The response is:
“There is no network connectivity between the test environment in
Bracknell to the live data centre in Belfast (or in 2008, in
Wigan/Bootle). So even if you were an IT wiz, you wouldn’t be able
to connect to the production service as there is no network to allow
it. Security and penetration testing against both Horizon and HNG
has been performed against the production environment to ensure
this is the case”.
42. In addition, in relation to a query about the import of an old Horizon Operating
Manual which apparently contained the statement “finance teams can no longer
‘* WBON0000344.
12 POL00130311; POL00188299; POL00098619. I was also provided with the questions asked by
Second Sight which had ‘driven the initial [Spot Review 5] submission’: WBON0000739.
13 POL00098619.
14 POL00031348.
Page 28 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
adjust client accounts on site”, Gareth Jenkins provided the following response
which was forwarded to me on 20 June 2013:
“There was never any capability for POL Staff to manipulate the Branch
accounts through Horizon. I think the Ops Manual is badly written.
I do remember the introduction of TCs in 2006 (I was the Architect
responsible for this as part of the IMPACT programme). What used to
happen before that is that the Branch was sent a piece of paper called
an Error Notice. This would then instruct them to carry out some specific
transaction at the Branch. These were often ignored. The whole point of
TCs was to simplify and speed up the process and enforce conformance.
There may also have been a mechanism by which POL could manipulate
the branch accounts in their old accounting system (CBDB — owned and
operated by POL or CSC on their behalf), but I never had any real
understanding of that system.
Therefore I think this is yet another red herring!”®
43. This was subsequently confirmed in the same email chain by a POL employee,
Rod Ismay, as follows: “As regards the words “...finance teams can no longer
adjust client accounts on site...” — “On site” meant “on site in Chesterfield P&BA”
not “on site in branch”.”"®
44. Rod Ismay’s email also answered (in bold) various other questions posed by
Second Sight, as follows:
“1. whether, before December 2006, any POL employees were able to
input transactions directly into branch accounts... and if so,,,
No — not into a branches accounts. See next email
*5 POL00098619.
16 POLO0098619.
Page 29 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
2. whether - and when and how - SPMRs/Branch staff were informed
whenever such interventions occurred
N/A - See next email
3. where POL staff having that capability were based
No such capability
4. what transaction types were involved and, lastly...
No such capability
5. what User IDs were applied to the transactions so executed.
No such capability”
45. From my emails it appears that Steve Allchorn prepared a first draft of POL's
revised response to Spot Review 5,'° which I then reviewed and made minor
comments on." The final version of the reworked response was sent to Second
Sight on 21 June 2013” and included the following information additional to that
provided in the original version:
“Summary
An assertion has been made by Mr Michael Rudkin that during a visit to
the Fujitsu Bracknell site on Tuesday 19th August 2008, he observed an
individual based in the basement of the building who demonstrated the
ability to access ‘live’ branch data and directly adjust transactions on the
Horizon system.
Given the amount of time that has passed, neither POL nor Fujitsu have
any record of Mr Rudkin attending the Bracknell site.
It has however been determined that the basement of Fujistu's building
contained a Horizon test environment. This environment was not
17 POL00098619. The further email to which Rod Ismay refers was an email to Second Sight,
which was subsequently forwarded to me by Steve Allchorn: POL00031350.
18 WBON0000737.
18 WBON0000361; WBON0000363.
20 WBONO000389.
Page 30 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
physically connected to the live Horizon environment. It was therefore
impossible for anyone in this room to have adjusted any live transaction
records though they may have shown Mr Rudkin some form of
adjustment to the test environment.
Test environment only
The key point here is the phrase ‘test environment’. In August 2008, the
live Horizon Data Centre was dual-located in Wigan and Bootle. Access
to the live site was strictly controlled and one could not interfere with the
live transaction databases from the test environments at Bracknell.
To create the test environment at Bracknell, POL/Fujitsu physically built
a completely separate set of servers that reflected the live configuration
in Wigan/Bootle. These servers were hosted in the basement in
Bracknell, along with test counters to connect to them. Access to the test
environments then (and which remains the case now) was controlled via
secure rooms and user logon authentication.
Critically, there was no physical connection between the live and test
environments. The test environments at Bracknell could not access nor
manipulate any data in the live environment.
However, as a test environment, there would have been terminals where
interrogation of the test copies of the live databases would have been
possible. To a lay person, this may look like activity in the live
environment. But, to be clear, this would have been interrogation of the
test databases only, as there was complete physical and technological
separation between the test and live systems.”2'
46. The response then answered Second Sight’s seven original questions in
substantially the same terms as previously with only minor tweaks (as POL’s
further investigations had not changed the substance of these answers), and
21 POL00243412.
Page 31 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
continued (under the heading “What access was historically available to live
Horizon data?’):
“As referenced in the Spot Review, the Horizon Operating Manual from
2006 notes that the introduction of a new system meant that POL could
"no longer adjust client accounts on site". POL has been asked to clarify
whether this meant that POL could access and change live Horizon data.
In parallel to Horizon, POL operates a finance IT system. This finance
system manages the relationships between POL and its product
suppliers. These relationships are the "client accounts" referred to in the
Operating Manual.
In 2006, POL upgraded its finance system to a new SAP finance system.
Before this upgrade, transaction records were sent from Horizon to the
old finance system. When certain types of error were made in recording
transactions in branch, POL's Product and Branch Accounting (P&BA)
team based in Chesterfield could make manual adjustments to the
finance system records so that the client accounts would be corrected.
This is what the Operations Manual meant by an adjustment being "on
site" — the site being the Chestefield site of the P&BA team.
For clarity, the manual adjustments to the finance system did not change
the Horizon records and therefore did not change the branch's local
accounting position.
Post 2006 and the introduction of the SAP system, POL changed this
process. The errors that would historically have been corrected by
manual adjustment are now corrected by way of a transaction correction
being issued to the branch. On the SPMR accepting the transaction
correction, the Horizon data is updated and this flows through to the SAP
finance system.
It was this change of process that led to the above entry in the
Operations Manual. This change, and the ability to access the old
Page 32 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
financial system, is also entirely unrelated to the test environment at
Bracknell.’22
47. The assistance I provided in preparing this updated response was — as with the
conventional responses to Spot Reviews — fundamentally a drafting exercise
involving the collation of information which had been gathered from within POL
and from Fujitsu.
48. At the time the reworked response was provided to Second Sight, the POL
employee with whom Michael Rudkin had allegedly met had still not been
identified. On 1 July 2013, Rodric Williams forwarded an email from Second Sight
identifying Martin Rolfe (a Senior Test Analyst in POL’s Bracknell-based IT team)
as the individual in question. Second Sight had asked POL “to get Martin’s side
of the story straight away’, so Rodric Williams instructed WBD to obtain a witness
statement from him.?° I cannot remember precisely how the decision was made
to take a witness statement (as opposed to capturing the information in some
other form); I recall thinking that a witness statement would be the appropriate
vehicle to capture Martin Rolfe’s account, and that this would have more weight
than merely providing a written update on behalf of POL based on the information
he provided.
49. I tasked Andrew Pheasant, an associate in the firm, to carry out this work and
had little involvement personally. I cannot recall whether I reviewed and
commented on the statement once drafted, but I am aware that Martin Rolfe
confirmed that the test environment in the basement of the Bracknell office had
no ability to connect to the live Horizon system — in other words, the information
22 POL00243412.
23 WBON0000743; POL00296872.
Page 33 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
he provided was consistent with the response to Spot Review 5 that had been
provided to Second Sight. I can see from my emails that I supplied the final
version of Martin Rolfe’s statement to Simon Baker on 1 August 2013, who
forwarded it to Second Sight the same day.”4
50. For completeness, I note that following receipt of the reworked response to Spot
Review 5, Second Sight continued to query the assertion that “/t/he POL
Bracknell Team had no access to the live system”, on the basis of certain emails
supplied by POL which stated that “although it is rarely done it is possible to
journal from branch cash accounts. There are possible P&BA concerns about
how this would be perceived and how disputes would be resolved”.#” Second
Sight ultimately reported that it was told by POL that this comment “describes a
method of altering cash balances in the back-office accounting system, not
Horizon” and that “none of the POL employees working in Bracknell in 2008 had
access to the back-office accounting system’.*® I do not believe I had any
involvement in the provision of this follow-up information, which would have been
a matter between POL and Second Sight following submission of the Spot
Review response.
E. SECOND SIGHT INTERIM REPORT AND ESTABLISHMENT OF THE
MEDIATION SCHEME (Q16 to Q19)
(i) POL’s preparation for receipt of the Interim Report (Q16 and Q18)
24 WBONO0000919.
25 WBON0000389, forwarded to me by Rosie Gaisford on 5 July 2013: WBON0000366.
26 Interim Report, §§1.12-1.13: POL00099063.
Page 34 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
51. On or around 27 June 2013 I received a call from Rodric Williams. The
background was the investigation into Horizon being conducted for POL by
forensic accountants from the firm Second Sight. I do not recall what we
discussed on this call, but Rodric Williams followed up with an email on the same
date which explained that (i) Second Sight would shortly be presenting its Interim
Report to MPs and (ii) POL wanted to ensure that “all concerned [were] aware of
the responses [POL had given] to the issues raised” during Second Sight's
investigations.2” He asked me to provide a summary of POL’s responses to four
specific Spot Reviews (which were those that would ultimately be discussed in
the Interim Report), and to prepare a document setting out Second Sight’s
obligations to consider the evidence submitted to it by POL in the course of its
investigation.
52. To assist me in this task, Rodric Williams provided a document entitled ‘Raising
Concerns with Horizon’,?® which I understood had been agreed between POL,
JFSA and Second Sight the previous year (prior to my involvement), and which
set out (i) the process by which SPMs could raise concerns for consideration by
Second Sight, and (ii) the terms of reference for Second Sight's investigation into
those matters.
53. As requested, later that same day I sent Rodric Williams a table setting out the
key issues raised by the 10 Spot Reviews which POL had asked us to assist with
their responses to, and a one-line summary of POL’s position on each.”
27 POL00021822.
28 POL00021823; POL00021824.
2° WBONO000741.
Page 35 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Separately, I provided the precis of Second Sight’s obligations with respect to
considering POL’s responses and supporting evidence.°°
54. I surmise (though am speculating based on reviewing the above emails now) that
this request was part of wider activity by POL to draw together relevant
information so as to be in a position to consider and respond to Second Sight's
Interim Report when it came in. For example, on 28 June 2013 I was asked by
Rodric Williams to pull together a summary of the effect of the so-called
“Suspense Account bug”™" and the steps POL had taken in relation to it.°? As
directed by Rodric Williams I liaised with Andrew Winn (a Relationship Manager
in POL’s Financial Service Centre) to prepare a summary® and provided the text
of this to Rodric Williams later the same day. I can see references to a ‘briefing
paper’ in these emails, which the Suspense Account bug summary was
apparently fed into. My knowledge of the Suspense Account bug at this time is
dealt with further below at §268.
55. As another example, on 30 June 2013 Rodric Williams emailed Gavin Matthews
asking WBD to establish “whether bugs [in] the Horizon system [had] ever been
specifically discussed in any proceedings”, including in shortfall claims.°> On 1
July 2013 I reverted to Rodric Williams in the following terms:
“Other than Castleton and Misra, we are not aware of any litigation that
has involved an allegation of an actual bug in Horizon.
3° WBON0000364; WBON0000365.
31 This is what it came to be termed in the group litigation and I adopt that nomenclature here. At
this point in time in 2013, it was sometimes referred to colloquially within POL as the “14 bug”
or similar, because it was understood to have impacted 14 branches.
32 WBONO0000742.
33 POL00341337.
34 POL00407493.
38 WBONO000131.
Page 36 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
However, this is based on anecdotal discussions inside Bond Dickinson.
Please bear in mind that we have handled 100s of cases over the last 5-
10 years for POL so (absent a case by case review) it's impossible to
say for certain that no SPMR alleged a Horizon bug.
We are however confident that no case in the last 2 years has involved
an allegation that there is a specific flaw in Horizon. There are anumber
of cases handled by our paralegal team that have been put on hold
because an SPMR has alleged problems with Horizon. These cases are
suspended pending the Second Sight report. It may be that on closer
inspection these cases reveal a specific complaint about an error in
Horizon however we would need to undertake a deeper review of each
case to determine this.’°
56. WBD had acted for POL in the civil proceedings with Mr Castleton and I was
aware from discussions with Rodric Williams that Mrs Misra had been prosecuted
by POL and that the reliability of Horizon had been considered during the trial
(WBD was not involved in that case). As part of the investigations required to
respond to the email, I had spoken to the lawyer (I do not recall this conversation
but suspect this was Stephen Dilley) who handled the Castleton case. I explained
that Mr Castleton had been unrepresented by the time the case got to trial, so
there had not been a sustained analysis of alleged defects in Horizon in his case.
However, brief reference had been made to a known bug in Old Horizon, the
“Callendar Square bug” (also sometimes called the “Falkirk bug”, so-named
because it had affected the Callendar Square branch in Falkirk), in the following
way: in his cross-examination of Anne Chambers, the Fujitsu employee who gave
evidence about Horizon in the case, Mr Castleton had described “complaints
from another branch, which he did not identify [but which Ms Chambers]
36 POL00407496.
Page 37 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
immediately recognized ... with confidence as being a branch at Callender
Square in Falkirk” (this part of my email is an excerpt from the High Court's
judgment). Ms Chambers had acknowledged the Callendar Square bug but
stated that there was nothing to suggest it had affected Mr Castleton's branch,
and the Court accepted this evidence. I stated that I was still making enquiries
about the Misra case and would report back in due course.°”
57. Shortly afterwards I emailed Rodric Williams again identifying that Misra had
been a criminal case. WBD had therefore not been involved in these proceedings
but had subsequently “reviewed the transcripts ... to identify anything relevant to
Horizon”. I do not believe that I myself read the transcripts, but I did review notes
on them which were made by others. In Misra there had been a sustained
examination of Horizon, but on reviewing WBD's notes on the transcripts “/
[couldn't] see anything that look[ed] like” the Suspense Account bug, the
Callendar Square bug or the “Receipts and Payments Mismatch bug”® (being
the other bug that was at that time known to have occurred within Horizon).°? In
fact, the Callendar Square bug was referred to in the Misra case, although it was
not thought to have affected Ms Misra’s branch. I would have read WBD’s notes
of the transcripts only briefly before sending this email, and it seems I simply did
not pick up on this reference at the time. As with the Suspense Account bug, I
deal with my knowledge of the Receipts and Payments Mismatch bug and the
Callendar Square bug at this point in time further below, at §267 and §269.
Suffice it to say for present purposes that these three bugs were the only ones I
37 POL00407496.
38 WBON0000746.
3° This is what it came to be termed in the group litigation and I adopt that nomenclature in this
statement. As my email to Rodric Williams shows, at this point in 2013, this bug was also
referred to colloquially as the “62 bug” or similar, because it had affected 62 branches.
Page 38 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
was aware of at this time (and for a considerable period thereafter, until during
the group litigation). They all featured in Second Sight’s Interim Report because
details of them had been disclosed by POL to Second Sight (though I had no
involvement in this process).
58. Other work I undertook around this time, when POL was anticipating receipt of
the Interim Report, included reviewing a draft letter which was ultimately sent to
James Arbuthnot MP. My emails indicate that I was simply asked (by Rodric
Williams, on 3 July 2013) to check whether a section on “access to live data” was
consistent with POL’s response to Spot Review 5; I am not sure I was even made
aware who the letter was to be from.*° I provided minor suggested amendments
on 4 July 2013.41
59. In other words, in the period leading up to receipt of the Interim Report my work
for POL was in the nature of pulling together small pockets of information in
response to specific requests from POL, reviewing documents, and carrying out
drafting work. I was not engaged to provide advice on POL’s general strategy for
preparing for and responding to the Interim Report. The nature and level of my
involvement did increase substantially following the release of the report, as I
explain in the subsections that follow.
(ii) Views upon receiving the Interim Report (Q17)
60. On 4 July 2013, Rodric Williams emailed me to forewarn me that the report was
expected to arrive in draft at 10:30 the following morning, and asking me to be in
4° WBON0000757; POLO0190547.
41 WBON0000135; WBON0000136.
Page 39 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
a position to quickly turn around comments.*? I do not recall this email but my
understanding of it is that POL wanted me to highlight any key deficiencies in the
report and to identify any errors (particularly with respect to the Spot Reviews
and POL’s responses to them) so that these could be corrected prior to
publication. POL was only afforded a limited amount of time to provide
comments, so it was a case of quickly recording my initial reactions and spotting
obvious errors.
61. On 5 July 2013, Rodric Williams sent me the draft Interim Report.4? Shortly
afterwards, he sent me an older version of the draft report and asked me to
produce a compare version for him to review.“4 To the best of my recollection and
based on the correspondence I have reviewed for the purpose of preparing this
statement, this is the first time that I received both documents. It is apparent that
I provided the compare version as requested on 5 July 2013,° although I have
not been able to locate the original email in which I did this (or the comparison
document itself).
62. I provided my comments on the draft report on the same date.*® I have limited
independent recollection of what I thought about the report when I first read it,
but the comments I provided to Rodric Williams would have accurately reflected
my initial reactions. My headline views were that:
62.1. Second Sight had identified “no evidence of system wide (systemic)
problems with the Horizon software” (§12.2(a) of the draft report).4”
42 WBONO0000759
43 POL00021745.
44 WBONO000760.
45 WBONO000762.
46 WWBONO0000134.
47 Later paragraph 8.2(a) of the Interim Report: POL00099063.
Page 40 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
62.2. They had identified no gaps in POL’s responses to the Spot Reviews.
62.3. They had failed to undertake any real analysis or evaluation of SPMs’
complaints and POL’s responses, by and large just reciting what each side
had said. Against that background, they had expressed views and
conclusions without providing supporting evidence or even any real
reasoning. My comments reflect (and I recall thinking) that I found this
concerning, because: (i) Second Sight had been appointed for the express
purpose of providing a reasoned expert opinion which was supported by
evidence; and (ii) they had by this point been conducting their investigation
for just over a year.
63. Rodric Williams subsequently informed me that he had relayed my comments to
Second Sight (though I do not know to what extent he did so, or in what form
they were provided).*® On 7 July 2013, he sent me the finalised Interim Report
(POL00099063).*°
(iii) Establishment of the Mediation Scheme (Q19)
64. After the Interim Report was published, I recall that POL was concerned to find a
way to progress the remainder of the Second Sight investigation. I recall that
POL was dissatisfied with how little progress had been made by Second Sight
and the limited number of reasoned and evidenced conclusions which the Interim
Report was able to draw. I recall that Susan Crichton and Rodric Williams were
of the view that a general investigation into the Horizon system, including the
training and support provided by POL, was simply too big a task for Second Sight
48 WBONO000762.
49 WBONO000763.
Page 41 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
to manage. POL therefore wanted to shift the focus to something more
manageable and that could be completed within a reasonable time frame.
65. I understood that there was also an element of political pressure to conclude the
investigation (and Second Sight's apparent lack of progress was adding to this)
but I was not involved in the discussions with MPs and so gained this
understanding only second or third-hand from the POL legal team. I vaguely
recall that some form of Ministerial commitment or commitment to MPs was given
by POL that the remainder of the investigation would be conducted quickly, but I
cannot remember the details of this or may not have even been told the details.
66. My perception was that there was a genuine desire on the part of POL to get to
the bottom of the issues that had been identified by Second Sight, to find closure
for the SPMs who had raised concerns so far as possible, and to make
improvements to its processes if necessary.
67. In view of these factors (and, I dimly recall, because key members of POL’s in-
house legal team were either on annual leave or due to go on leave), within a
few days after publication of the Interim Report Susan Crichton turned to WBD
(including me) for advice and support in developing proposals for a way forward.
My emails suggest that we spoke by telephone on 11 July 2013 (though I have
no recollection of this conversation) and the following day she emailed me
floating the idea of an arbitration or mediation process as the means by which
POL could seek to directly resolve at least some of the individual SPMs’ cases:
“I have been giving some thought to how we might 'sort' what JFSA calls the
Page 42 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
‘toxic' cases.©° So a couple of things firstly given that this is about reaching a
conclusion I wondered if we could use either an arbitrator or a mediator - both
would be independent tho [sic] paid by POL, they could even sit between POL
and SS?".5!
68. I responded later that day with a brief explanation of the pros and cons of
mediation versus arbitration in this context:
“Had a quick chat with Gavin.
Arbitration will probably end up as formal and long winded as court
proceedings. We'd also lose a degree control - the process and timing
would be controlled by the arbitrator. I'm not attracted to this.
Mediation is a definite possibility. I could envisage a mediation between
POL and each SPMR (with also SS in the room - and perhaps
Shoosmiths?). This gives each SPMR the opportunity to voice their
views and discuss SS' findings. Having a mediator in the room would
help equalise the imbalance of power. Mediation would not commit POL
to any outcome (unless one was agreed by both parties) and could be
conducted on our timetable. If the mediations were run after SS's final
report, this may help ensure that the report focuses on general themes
whilst leaving specific cases to be heard in the subsequent mediation
process.
The risk is that mediation is usually set up with a view to reaching an
[sic] resolution. As discussed yesterday I doubt we will ever reach
closure on these cases. POL's comms team would therefore need a
robust media strategy to explain why the mediations will, in the majority
of cases, fail to reach a consensus between POL and the SPMR.
Otherwise, this may be spun as a failure to close out this matter.
50 “Toxic cases” was a term sometimes used by those involved (and I believe coined by the JFSA)
to refer to those cases which were (or were likely to be) particularly difficult to resolve as they
were particularly long-running or had attracted an unusually high level of media attention.
51 WBONO000766.
Page 43 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Cost: for a decent mediator (ie. corporate background / someone who
isn't going to roll over on hearing a sob story), we'd be looking at around
£1,000 - £3,000 per half day, though we should get a "bulk buy" discount!
I know a number of mediators who would be suitable for this project.”®?
69. In this email, I was trying to give succinct advice to set up a realistic mediation
scheme, recognising the practical reality (as I understood it) that many cases
would not resolve through mediation. I wanted the process to be genuinely fair
and accessible, thus my concern that SPMs should have a forum which gave
them a ‘voice’ and that steps should be taken (including potentially sourcing
independent legal representation for them) in order to ‘equalise the imbalance of
power’ between them and POL. I thought it important that if the mediation route
was adopted, a high-quality mediator should be sourced who would deal with
SPMs' cases in a way that was robust, even-handed, and fair to both parties.
70. From my emails, over the next 10 days or so I, others at WBD, and employees
of POL discussed different options for attempting to resolve individual SPMs’
cases by email, telephone and in meetings. For example:
70.1. On 12 July 2013, Susan Crichton forwarded me a proposal that had
apparently been made by JFSA to Paula Vennells (POL CEO).5° My
response indicates that I thought that the proposal was not dissimilar to
whatever option we had discussed by telephone the previous day (see
above, §67), albeit with some important (and potentially problematic)
differences. For instance, it appeared to: assume there would be a cash
settlement in all cases; imply Second Sight should have a role in
52 WBON0000767. Rodric Williams’ response to Susan Crichton's email, to which he refers, is:
POL00230639.
53 POL00407537.
Page 44 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
quantifying claims; and suggest that criminal convictions could be
addressed through the scheme.*4 Mr Flemington similarly expressed
concern about the suggested inclusion of convicted SPMs in JFSA’s
proposed scheme, pointing out the potential for inconsistencies between
the outcome of the scheme and decisions of the criminal courts.®®
70.2. On 17 July 2013, Susan Crichton emailed Gavin Matthews (with me in
copy) querying the possibility of independent adjudication. Gavin
responded setting out how an adjudication process might work, its
positives and negatives, ultimately recommending mediation as a more
appropriate model for attempting to reach resolution in individual SPMs’
cases in the short-term.°
70.3. On 19 July 2013, Mark Davies (Communications Director at POL) mooted
a proposal which would involve POL creating an independent panel
chaired by a QC, former MP or perhaps a Peer to hear evidence in
individual cases, and allocating funding to compensate SPMs in cases
where it was found that POL had failed to provide adequate training and
support.5” Susan Crichton emailed me the same day providing some
further background to this suggestion and seeking my views.*?
71. The foregoing is necessarily just a flavour of the relevant discussions that were
going on at the time. As they demonstrate, there were a range of complex and
54 WBON0000768. As Susan Crichton’s reply at the top of the chain shows, at the time I appear to
have mistakenly thought that this proposal emanated from Second Sight rather than JFSA, and
my comments should be read subject to that. Notwithstanding this, I believe I would have still
considered that the substance of the concerns highlighted in my email were valid.
55 WBONO000769.
58 POL00407548.
57 WBONO000775.
58 WBONO000776.
Page 45 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
competing considerations in play, including (but not limited to): the commitments
POL had given publicly following publication of the Interim Report; the
preferences of the POL Board; the relative cost and speed of the dispute
resolution model chosen; the design of that process and the need to ensure
fairness to both sides; the need to manage SPMs' and the JFSA’s expectations
(for example, because it was unlikely that it would be appropriate to offer a cash
settlement in all cases); the different circumstances of different cases; whether
and how to accommodate SPMs with criminal convictions; and how Second Sight
should fit into the overall structure of the process, given that (unusually) the
process would be intended to progress the resolution of cases that Second Sight
had started to investigate, in circumstances where Second Sight was to remain
engaged but there were legitimate concerns about the speed and efficacy of its
review to date.
72. \Iand my colleagues at WBD favoured mediation. A compensation scheme pre-
supposed that all complaints were well-founded (and this was not accepted by
POL), and an adjudication or arbitration scheme would take too much time to set
up and was therefore not apt to achieve swift resolution of the SPM cases which
were the intended object of the process. I also believed that an adjudication or
arbitration scheme would not be satisfactory because not all of the SPMs who
had come forward were expected to want (only) compensation. For example, I
anticipated that some would want reinstatement if they had lost their positions,
and others may have wanted an explanation or an apology, or a commitment that
POL would improve its performance in the future. I thought that mediation was a
more flexible process which would accommodate a wider range of possible
Page 46 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
outcomes, whilst ensuring an appropriate degree of formality and fairness to both
parties.
73. Against this background, I prepared a note (POL00117035) setting out some of
the advantages of mediation in the present context and outlining a possible
process. I provided this to Susan Crichton and others on 19 July 2013 together
with a breakdown of the estimated costs of mediation (see POL00117034). In
summary, the process envisaged was that there would be an investigation
involving Second Sight preceding any formal mediation (tying in with its ongoing
role to produce a report into the common themes it had identified as arising out
of the cases it was and would continue to review). The original intent set out in
the note was that, once Second Sight had produced its thematic report (which at
that time I understood to be due in October 2013),°° the product of the
investigative process would be used in any subsequent mediations together with
any findings by Second Sight. This was to ensure that (i) cases were investigated
prior to the mediation and (ii) POL would have sufficient information to be able to
address the relevant SPM’s complaints during the mediation. I anticipated that
mediations would generally be focused on breach of contract claims for civil
compensation with loss of earnings and/or repayment of monies received by POL
as the two main heads of loss (as well as being able to accommodate SPMs who
were looking for an apology or other non-monetary remedy).®
74. \t appears from POL00117034 that Susan Crichton discussed this proposal with
Paula Vennells and they felt that mediation should be available once Second
Sight had produced a case review, as opposed to waiting for Second Sight’s
58 POL00192226.
6° Cf. POLO0099445.
Page 47 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
thematic report. This was (at least in part) because Second Sight had objected
to producing a thematic report by October 2013. Second Sight were apparently
broadly in favour of this way forward, so Susan Crichton asked me to consider
the implications of the proposed change to the process I had previously outlined,
and for a discussion about possible terms of reference for a mediator.
75. Atthis time, the JSFA was being supported and advised by Kay Linnell who was
an accredited arbitrator and mediator. I recall meeting with JFSA, Second Sight
and POL to discuss the mediation proposal on several occasions over the
Summer of 2013 and I recall that Ms Linnell was broadly in favour. As such, it
appeared to me that a mediation scheme was also preferred by the JSFA.
Discussions continued within POL, and between POL, JFSA and Second Sight
in the course of July and August 2013 (some of which I would have been sighted
on and others not). The terms of the Mediation Scheme were in due course
agreed with the scheme opening to applications on 27 August 2013, and closing
on 18 November 2013.5"
76. For completeness, I note that during this period I also advised POL on the
structure of the Working Group to oversee the Mediation Scheme (which
organically grew out of the meetings between POL, JFSA and Second Sight over
the Summer of 2013) and the appointment of an independent third party to chair
the Working Group. These aspects of the scheme arose in part out of the JFSA’s
concerns that the investigative process and mediation stage should have
independent oversight. This led to (i) the creation of a Working Group comprising
POL, Second Sight and JSFA pursuant to agreed terms of reference,® and (ii)
®1 WBON0000778; WBON0000784; WBON0000787; WBON0000790.
2 WBON0000817.
Page 48 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
the appointment of Sir Anthony Hooper, a retired High Court Judge, as the
independent chair pursuant to agreed terms of reference.® Sir Anthony took up
his appointment as Chairman with effect from 29 October 2013.
77. I deal with the Inquiry’s questions concerning the operation of the Mediation
Scheme from this time until my work on this aspect reduced in mid-2015 below,
in Section G.
F. REVIEW OF CRIMINAL CONVICTIONS AND RELATED MATTERS (Q20 to
Q31)
(i) Overview and nature of my role
78. In the course of 2013, whilst the work described above was ongoing, POL was
separately considering and taking advice from its criminal lawyers on what steps
it ought to take in respect of criminal proceedings against SPMs suspected of
theft, false accounting, and similar offences (where it had been the prosecutor).
I later became aware that POL was specifically considering what if any
disclosures needed to be made to SPMs who had previously been convicted of
such offences, in circumstances where the prosecution had relied (at least in
part) on Horizon data. This was because of advice it had received from its
criminal solicitors (not WBD) about a deficiency in Gareth Jenkins’ evidence
which consisted of his failure to reveal the existence of bugs in cases where the
integrity of Horizon Online had been in issue.
3 WBONO000789.
Page 49 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
79. As I explain further below, I had limited visibility of this work because it was
principally managed by POL’s external criminal solicitors, Cartwright King
Solicitors) “Cartwright King”, who had conducted recent prosecutions of SPMs
on POL’s behalf, together with POL’s in-house legal team (especially Jarnail
Singh who was then POL’s internal criminal law specialist). I had never — and
have never — had conduct of criminal proceedings against an SPM or indeed in
any criminal prosecution. The fact that I was involved at all is explained by the
following matters:
79.1. The civil (Mediation Scheme and later the group litigation) and criminal
workstreams were running at the same time and concerned some
overlapping issues (particularly as to Horizon) and, in some respects, the
same SPMs.
79.2. Cartwright King had been and continued to be POL’s criminal law advisers,
but as such they were involved with POL’s historic private prosecutions.
POL therefore wished to obtain separate criminal law advice at arm's
length from Cartwright King and (as I explain further below) WBD played
a role in facilitating that.
80. In view of these matters, the role I played in relation to the criminal law
workstream was a supporting one (for example, my firm acting as Brian Altman
QC’'s instructing solicitors); or it was incidental to my position in respect of the
civil matters which was my primary remit.
4 For example, POL sought a new expert to replace Gareth Jenkins. Cartwright King led on the
search for that expert and it was their role to advise POL on the suitability of prospective experts
to give evidence in criminal proceedings. However, since POL thought that the new expert may
be able in due course to give evidence in any civil proceedings involving a challenge to Horizon,
I was sighted on the conduct of the search and in one or two instances sought proposals from
prospective experts on POL's behalf: see for example WBON0000795. The proposal received
Page 50 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
81. In order to maintain the chronology of this statement, in this section I primarily
focus on answering Q20 to Q31 of the Request insofar as those questions relate
to the work which I undertook in 2013 and early 2014. Subsequently, in early
2015, the CCRC opened a formal investigation into the convictions of certain
SPMs (and the scope of that investigation later expanded to consider other
SPMs'’ cases); and Brian Altman QC undertook a further review in 2016. I will
deal with my limited involvement in those matters below (see especially §§229-
232, §§296-297, §§458-465, §§596-598, and fn. 250).
(ii) The Helen Rose Report (Q20) and review of criminal convictions
82. Based on my email records, I first received the Helen Rose Report concerning
the Lepton SPSO (POL00022598) from Rodric Williams on 3 July 2013. Rodric
Williams forwarded me an email from Dave Posnett of POL (dated 14 June 2013)
which had the report attached and asked that I take a look at it so that I could
give him an overview on the phone at some point that day. Dave Posnett’s email
recorded that the report had been produced by Helen Rose, and that it
“concern[ed] a ‘system reversal’ of a transaction following a system failure”. For
context, this was a reference to the automatic transaction reversal process,
which was a safeguard in Horizon Online that activated if a terminal in a branch
lost power or the telecoms line was disconnected. The automatic reversal of
transactions in these circumstances was (as Dave Posnett’s email put it) “normal
practice”, but Dave Posnett appears to have been concerned that the relevant
data logs gave a misleading impression that such automated reversals had been
from Deloitte for this work is at WBON0000773; ultimately, nothing flowed from this proposal
and it was unconnected with Deloitte's later work (addressed later).
®5 WBONO0000751.
Page 51 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
entered manually by the SPM. Dave Posnett appears to have had in mind that,
if the SPM did not appreciate that a reversal had taken place and did not take
steps to match this physically in the branch (e.g. by handing back any money
that had been taken from the customer), a discrepancy could result, and that it
may not be apparent from the data logs what had happened.
83. I do not recall Rodric Williams’ email of 3 July 2013 asking me to give him an
overview of the Helen Rose Report or what I did in response to his request. Nor
can I find any emails that shed further light on this. I cannot recall what my initial
views on the contents of the report were, nor what impression of its contents I
gave to Rodric Williams (if I did in fact call him to provide an overview, which I
cannot now remember).
84. However, and on the basis of my present knowledge, I make the following
observations as to what my initial views of the Helen Rose Report’s contents
might have been:
84.1. First, the only aspect of the report that I would have considered relevant
to my role would have been the criticisms of the automatic transaction
reversal process. At the date I received the report, I would likely have
appreciated that WBD had already considered the mechanics of the
transaction reversal process by reference to what had taken place on 4
October 2012 at the Lepton SPSO, as this was the subject of Spot Review
4,66
66 In this regard, I note that Dave Posnett’s email of 14 June 2013 was incorrect in stating that the
Lepton SPSO did not “feature as part of [the] 2" Sight Spot Reviews”: WBON0001725.
Page 52 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
84.2. Second, I would have recognised this as Spot Review 1 because I had
personally been involved in redrafting POL’s response to Spot Review 1 in
April 2013.57
84.3. Third, I would have recalled that POL’s response to Spot Review 1 had
concluded that (i) no failing in Horizon had been demonstrated in the sense
that the system had operated as intended in accordance with its design,
and (ii) the system provided adequate notification of automatic transaction
reversals that occurred when it was unable to connect to the Data
Centre.®* With this in mind, I would have reviewed the criticisms of the
transaction reversal process that featured in the Helen Rose Report, in
order to satisfy myself that they had already been addressed as part of
Spot Review 1 (and that no amendments or additions to POL’s response
were required).
84.4. Fourth, as to point (i), I would likely have thought that POL’s conclusion in
its Spot Review 1 response, that no failing in Horizon was demonstrated,
broadly aligned with Helen Rose’s summary that “the system ha[d]
behaved as it should and [she] did not see this scenario occurring regularly
and creating large losses’.
84.5. Fifth, as to point (ii), I would have noted that Helen Rose made a particular
criticism of the fact that the data logs readily available to POL did not
clearly differentiate between an explicit transaction reversal completed by
the postmaster, and an automatic transaction reversal as part of the
recovery process arising from connectivity issues (i.e. the point alluded to
7 POL00098035.
8 POL00186743.
Page 53 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
by Dave Posnett). In its Spot Review 1 response, POL had acknowledged
this point but concluded that it did not create a serious difficulty, as it was
possible to determine what had happened from the disconnect and
recovery receipts that would physically print in the branch at the relevant
time. Therefore, this criticism is unlikely to have surprised or particularly
concerned me at the time I received the Helen Rose Report, as the same
point was raised and addressed as part of Spot Review 1.
85. Save for the matters set out above I had no insight into the background to the
Helen Rose Report, and I do not believe I would have had any other views on its
content. I did not know who had commissioned it or why (beyond the bare fact of
the events of 4 October 2012 having happened).
86. From the advice note prepared by Cartwright King dated 15 July 2013 (the
“Clarke Advice’”),®? I can see that at some point POL passed the Helen Rose
report to Cartwright King. The report quoted an email to Gareth Jenkins which
read: "! know you are aware of all the horizon integrity issues”. Through the
inquiries described in the Clarke Advice, Cartwright King established that Gareth
Jenkins had been aware of two bugs in Horizon Online (the Receipts and
Payments Mismatch bug and the Suspense Account bug) at a time when he gave
evidence in criminal proceedings to the effect that Horizon Online was ‘bug-free’.
On this basis, Cartwright King advised POL that (i) Gareth Jenkins’ credibility as
an expert witness was called into question, and (ii) there may have been material
non-disclosure of the Suspense Account and Receipts and Payments Mismatch
bugs in some past and ongoing prosecutions concerning Horizon Online, which
6° POL00193002.
Page 54 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
would now need to be rectified. I was not involved in providing the Helen Rose
report to Cartwright King or in the seeking or preparation of the above advice.
87. The Clarke advice was sent to me (I believe) for the first time on 17 July 2013 by
Susan Crichton.”° From my emails it seems that POL may have alerted me to the
substance of Cartwright King’s thinking slightly before this, on or around 8 July
2013, though I do not recall this email or the conversation to which it alludes.”
At any rate, I have identified nothing that suggests that I was appraised of the
matters in the preceding paragraph when I first received the Helen Rose Report
on 3 July 2013. Nor did I have the means to work them out for myself, not having
been involved in any prosecutions of SPMs and not being aware of the evidence
that Gareth Jenkins had given in some of them.
88. Once I was made aware of the substance of the Clarke advice, I would have
understood that the points it raised meant there had been possible disclosure
failures in previous prosecutions. I would have taken Cartwright King’s opinion
on this issue and on the issue of what POL needed to do next to discharge its
duties as prosecutor at face value, this being Cartwright King’s area of
specialism.
89. I recall being aware of the fact that as a result of this development, Cartwright
King began undertaking a review of historic prosecution files to determine what
disclosures POL needed to make to comply with its prosecutorial duties. WBD
did not advise on this review and disclosure process, and nor were we sighted
on its progress or conduct, save that we played a limited role in arranging the
instruction of Brian Altman QC as described below. As part of that disclosure,
70 WBONO000770.
71 WBONO000765.
Page 55 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Cartwright King disclosed the Helen Rose Report to some convicted SPMs and
decided to redact parts of the report. WBD did not play any role in that disclosure
or the decision to redact that document before it was disclosed; I became aware
of this only later, when a request for disclosure of an unredacted version of the
report was made on behalf of certain SPMs who were participating in the
Mediation Scheme (see below, §§176 ff).
(iii) The CCRC’s July 2013 letter and the appointment of Brian Altman QC (Q29
and Q31)
90. Soon after the publication of Second Sight’s Interim Report, on 12 July 2013, the
CCRC wrote to POL seeking information about the number of SPMs who had
been convicted (following a guilty plea or unsuccessful appeal) in circumstances
“where evidence from the Horizon computer system [was] relevant’, and asking
what action POL was taking in such cases (POL00039996). This letter was not
initially referred by Susan Crichton to me but to Cartwright King.
91. On 16 July 2013, Susan Crichton emailed me (POL00039996) expressing
uneasiness as to advice POL had received from Cartwright King in relation to this
initial letter.’2 This was the point at which I was first made aware of the CCRC’s
letter. Susan Crichton described Cartwright King’s advice as “odd... as if given
on a take it or leave it basis’, by which I understood her to be referring to their
comment at the bottom of the advice (POL00039993) in relation to their
suggested draft response to the CCRC (“Please feel free to use it, or any part of
it, (or not) as you will...”). Susan Crichton commented that “somehow it feels as
72 POL00039996.
Page 56 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
if there is a conflict here [i.e. in Cartwright King’s position]’, which I took to be an
allusion to the fact that in light of Cartwright King’s prior involvement in POL’s
private prosecutions of SPMs, it might be inappropriate for them to direct POL’s
response to the CCRC in connection with a possible independent review of those
same prosecutions.
92. I cannot recall what if any discussion I had with Susan Crichton on the back of
this email, but I note that Gavin Matthews responded later that day saying that
he had taken an “initial look” (so it may be there was no need for me to pick it up
with her).’° Gavin agreed that he would expect a solicitor advising on a letter of
this nature to give a clearer steer as to how to respond, and offered to identify
some criminal barristers to assist. He also commented, albeit from a civil
practitioner’s perspective, that Cartwright King’s draft response was poorly
phrased in that it did not reflect the fact that the Interim Report “found there to be
no systemic problems with Horizon’.
93. My emails indicate that Gavin subsequently had some discussions with Susan
Crichton and other members of the POL legal team about the way forward. In
short, he recommended that an independent criminal QC be instructed to
oversee the review then being carried out by Cartwright King (“to check that their
tactical approach is now overseen by someone completely unbiasea”).’ This
was the approach ultimately taken by POL, and Brian Altman QC was appointed
to supervise Cartwright King’s work. Brian Altman QC’s remit also went beyond
this in that POL instructed him to advise on how to deal with any review by the
CCRC, and on its wider prosecution strategy going forwards.
73 POL00407546.
74 WBON0000133.
Page 57 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
94. I drafted a holding response to be sent to the CCRC whilst the practical detail of
the instruction was sorted out.’> Gavin thereafter liaised with Brian Altman QC to
settle a more substantive response.’® I note that Gavin’s email providing that
letter to Susan Crichton (with me in copy) anticipated that a follow-up letter would
need to be sent. My emails indicate that Cartwright King and I were sent a copy
of this later letter and invited to comment; though I do not believe I provided any
comments given that the draft had come from Brian Altman QC via WBD.” For
completeness, my recollection is that the CCRC did not become especially active
at this stage and essentially monitored the matter in the background until early
2015 when it opened a formal review.
95. Because part of the purpose of Brian Altman QC’s instruction was to
independently evaluate Cartwright King’s review of SPMs’ convictions, bearing
in mind the role they had played in the past prosecutions, it was considered
prudent for his instructions to not to come from Cartwright King. POL could have
instructed Brian Altman QC directly through its in-house legal team, but I recall
that POL's reason for asking WBD to do this on their behalf was so that we could
assist with administrative matters such as preparing bundles of documents for
Counsel and organising conferences. Given WBD's parallel work on the
mediation scheme arose also from Second Sight's work and reports, my firm was
the obvious port of call. Gavin took the lead on this instruction rather than me.
POL understood that I was not a criminal lawyer and that WBD was acting as a
professional conduit for instructions to Brian Altman QC.
75 WBONO000777.
76 WBONO0000782; POL00297983.
77 \WBONO001705.
Page 58 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
96. With respect to the Inquiry’s Q@29, WBD did not advise on or conduct any review
of past prosecutions of SPMs. Nor did I advise on POL’s duties of disclosure
towards convicted SPMs, either generally or in individual cases, this being
squarely within the remit and expertise of the criminal law specialists (Jarnail
Singh, Cartwright King, and Brian Altman QC). As and when I advised on issues
of disclosure it was solely from a civil law perspective. I was not instructed to
advise on such issues in the criminal law context and POL's in-house legal team
knew that I had no expertise in those matters. Where issues of criminal procedure
(including disclosure) happened to be relevant to my own work, I endeavoured
to establish what I needed to know from the criminal lawyers and deferred to their
advice.
97. This can be seen in an email which I sent to Cartwright King on 5 August 2013
to which Harry Bowyer (in-house counsel) responded the following day. At this
time, I was engaged in setting up the Mediation Scheme, which (it had been
determined) would be open to SPMs with historic convictions based on Horizon
data. Mr Bowyer's responses are shown in red below:’°
“Im helping POL set up a mediation scheme to address SPMRs
concerns about Horizon and have a couple of quick questions with a
criminal angle that I hope you may be able to help with. Apologies but
we need a response relatively swiftly - close of business today if
possible.
1. Privilege
The mediations will be confidential and subject to "without prejudice"
privilege. Some will involve SPMRs who have been prosecuted.
78 WBONO000806.
Page 59 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Essentially, we're trying to determine whether criminal procedure /
disclosure duties trump privilege.
Disclosure always trumps privilege. If we are in possession of material
that undermines our case or helps the case that the defence are trying
to put forward then we are obliged to disclose unless there is a basis for
a Public Interest Immunity application. There are ways of disclosing
evidence that do not disclose the way that it was obtained e.g. Section
10 admissions or disclosure notes
Will "without prejudice" privilege prevent a SPMR from repeating matters
discussed in the mediation in later criminal proceedings (ie. an appeal)?
No - once the information is out it cannot be put back in the box. If the
sub postmaster discovers something that undermined our case or would
have supported his then he can use that as the basis of an appeal
subject to the rules of evidence, admissibility etc.
Likewise, if something is said during a mediation that may be material to
a SPMR's conviction, is POL obliged to disclose that information to
Defence Counsel even though it may have been obtained during without
prejudice discussions?
If the material comes from the SPMR himself then the defence already
have the information - we do not run the defence case - we just have to
make sure that they have the material that enables them to run it - the
difficulty arises if the subpostmaster refers to something that had not
been canvassed by the defence which triggers further disclosure. It may
be that a remark made by the SPMR might open a new area of disclosure
that had not been considered by the disclosure officer because the
defence had not mentioned it.
[ol
3. Disclosure duty
We've prepared a document that will be going to SPMRs to explain the
mediation process. That document contains the statement below. Does
this accurately capture POL's disclosure duties and the appeal process?
Page 60 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
"What if my case involves a criminal prosecution or
conviction?
You may put your case through the Scheme even if you have
already received a Police caution or have been subject to a
criminal prosecution or conviction.
However, Post Office does not have the power to reverse or
overturn any criminal conviction - only the Criminal Courts have
this power.
If at any stage during the Scheme, new information comes to
light that might reasonably be considered capable of
undermining the case for a prosecution or of assisting the case
for the defence, Post Office has a duty to notify you and your
defence lawyers. You may then choose whether to use that new
information to appeal your conviction or sentence."
Accurate and succinct! - As stated above no one who's prosecution is
live should be in the scheme.”79
(iv) Horizon Regular Calls (Q21 to Q28)
Inception of the Horizon Regular Calls
98. To my knowledge, it was Cartwright King who advised POL to set up the Horizon
Regular Call in around July 2013. I do not believe anyone at WBD had a hand in
this. I later understood that Cartwright King recommended this step in order to
assist POL with its ongoing disclosure duties in light of the fact that Cartwright
King now considered that there had been material non-disclosure, in certain
criminal proceedings against SPMs, of bugs in Horizon that POL was aware of
79 The draft Mediation Scheme pack reflecting this advice is POL00145832, circulated on 6 August
2013: WBONO0000787.
Page 61 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(as reflected in the Clarke Advice).®° Their idea was to have a single forum in
which different departments within POL could share information about issues
raised by SPMs.
99. I do not now have a good recollection of these meetings but my firm’s file
indicates that I attended most of the calls between 19 July 2013 and 15 January
2014 personally, with a few gaps, and that I attended one further call on 19
February 2014. I do not believe I attended any subsequent calls. I believe the
reasons why I would have initially attended personally were as follows:
99.1. The focus of discussion at the Horizon Regular Call, namely technical
issues with Horizon, was of interest given the issues being considered by
Second Sight. Therefore, I wanted to hear what was said on the first few
Horizon Regular Calls and to see what matters were raised and how.
99.2. I could usefully feed in anything to do with Horizon coming from the
Second Sight side. During this time (namely, the second half of 2013)
Second Sight was transitioning from conducting Spot Reviews to
facilitating the Mediation Scheme.
99.3. I could advise as to any civil liability issues which might arise from any risk
to the past criminal prosecutions.
Horizon Regular Call on 19 July 2013
100. I am asked to specifically consider the note of the Horizon Regular Call on 19
July 2013 (POL00083932); that note records that I was present. I have no
recollection of this call although I have no reason to doubt that I was present. My
80 POL00193002.
Page 62 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
firm’s records show I received an email invitation on 18 July 2013 from Ben Thorp
(a WBD employee then on secondment to POL’s legal team), which noted that
Rosie Gaisford (another WBD solicitor) had spoken to me about the call.°’ I do
not recall the conversation with Rosie. I have no recollection of POL providing
me with any other briefing before the first call, but see that according to the note
Rob King opened the call with an explanation of the call's purpose including that
"No minutes circulated, but we will be taking notes".
101. There are two entries as part of this note that are attributed to me, and about
which I am asked.
102. First, Q21.5 refers to the record that I “Commented on need to limit public debate
on the Horizon issue as this may have a detrimental impact on future litigation”.
I do not recall making this comment or exactly what I meant by it. However,
reviewing this comment now I observe that the Second Sight report had been
published a little over a week previously and POL was concerned that media
reporting could stir up further challenges to Horizon in circumstances where it
believed that Horizon remained robust, including (potentially) in the form of
unsubstantiated civil claims. I observe from the note that a significant proportion
of this meeting appears to have involved discussion of these types of external
communications issues, e.g. the need to keep an eye on internet forums where
the Horizon system was being discussed (Rod Ismay), an article in the Telegraph
(Ruth Barker), and the need to make clear to SPMs in the wider business network
what steps POL were taking to address the issues in the Second Sight report
81 WBONO000772.
Page 63 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(Nick Beal). I believe the comments attributed to me would have been part of the
flow of that discussion.
103. Second, Q21.6 states that I “Spoke about emails, written comms, etc ... if it’s
produced it’s then available for disclosure, if it's not then technically it isn’t’.
Again, I do not recall making this comment.
104. I believe that my comment could only have been intended to make sure POL was
aware of the fact that electronic documents and other forms of written
communication would be caught by the rules of disclosure in civil proceedings. It
was my function to advise my clients on (for example) what constitutes a
document in the civil context (which includes electronic documents).
105. I do not understand the reference to ‘produced’ in the comment that is attributed
to me. The word ‘produced’ is not a word I would ordinarily use in this context. I
think of the word ‘produced’ in the context of running a ‘production’ of documents
from a data room (which is a technical e-discovery process), which are then given
to the other side in civil litigation as part of disclosure. I would not use this word
when describing the creation of new documents by my client which are then
potentially disclosable. The use of this word suggests to me that the note is not
an accurate record of precisely what was said.
106. To be clear, I would also not have purported or attempted to comment on POL’s
criminal law disclosure obligations. That role fell to Cartwright King, whose were
present at this meeting. If my advice was in any way unconventional or liable to
cause confusion as to POL’s criminal law duties, I have no doubt that Martin
Smith would have said something to correct the position (and from the minutes it
does not appear that he did).
Page 64 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Horizon Regular Calls on 24 July, 7 August and 14 August 2013
107. I am also asked to specifically consider the notes of Horizon Regular Calls that
took place on 24 July 2013 (POL00083933), 7 August 2013 (POL00083931) and
14 August 2013 (POL00083930). I am not listed as an attendee of the call on 7
August 2013 and I have no reason to believe that I attended it.
108. As to the other two calls, again I have no memory of the specifics of them. I
therefore do not recall any discussion of the need to take or not to take minutes
or notes of the Horizon Regular Calls; or of the format in which this should be
taken; or of the Helen Rose Report (which is also not referred to in the notes of
either call). I have however reviewed the note of the Horizon Regular call which
took place on 31 July 2013, and these do mention the report.®? I appear not to
have attended that call (and cannot remember attending it) so cannot comment
further on what was discussed.
Minute-taking and record-keeping in respect of the Horizon Regular Calls
109. My firm did not, at first, play a role in minute-taking at the Horizon Regular Calls.
It appears to have been POL who was responsible for, and took, the notes of the
four calls that took place between 19 July and 14 August 2013. These notes were
not prepared by WBD: they were sent to me for the first time as a batch by Dave
Posnett of POL on 16 August 2013 (POL000139691).®°
110. Based on the note of the first call on 19 July 2013 (POL00083892), I had
suggested during that meeting that all lists of cases should be sent to Rosie
Gaisford. Around that time, and as a separate exercise to the weekly calls, I recall
®2 POL00193767.
®3 POLO00193596.
Page 65 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
that Rosie had been placed on a short-term (c. 3-4 week) placement to POL with
the primary task of collating a list of all the past criminal prosecutions and civil
cases that POL had instigated. This was because POL did not hold a master list
of these cases. Against this context, I believe my suggestion during the 19 July
2013 call was aimed at ensuring that any known cases were fed through to
Rosie's master list. I do not believe this was a reference to Rosie keeping a list
of Horizon issues or a minute of the calls, nor do I remember Rosie ever keeping
such lists or minutes.
111. Later, in August 2013, WBD were asked to take on the role of keeping minutes
for the Horizon Regular Call. The background to this was as follows:
112. On 13 August 2013, I received a call from Susan Crichton and Hugh Flemington.
I have no independent recollection of this call but an email I sent to Gavin
Matthews and Simon Richardson of WBD states that it was because Cartwright
King had advised POL that it needed to “track and investigate every single
complaint, query or issue about Horizon in order to comply with criminal
disclosure duties”, and POL was concerned that this would be “very difficult, if
not impossible, for POL to achieve”.*4 I believe therefore that they called me to
discuss the practicalities of implementing the advice and to ask me to seek input
from Brian Altman QC on this point® (note, I would not have advised on the
correctness of Cartwright King’s advice, nor would I have been asked to).
113. Following that call Susan Crichton sent me an email® attaching the written
advice that Cartwright King had prepared on POL’s duty to record and retain
®4 WBONO001710.
®5 See §§123 ff below regarding the conference with Brian Altman QC on 9 September 2013.
86 WBONO0000791.
Page 66 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
material (including information) by virtue of their role as private prosecutor (the
“Second Clarke Advice”), saying that she thought I “might be interested to see
what had started that particular ‘hare’ running’.®”
114. In the Second Clarke Advice, Cartwright King raised various concerns about the
approach some within POL were allegedly taking to keeping records of the
Horizon Regular Calls:
“At some point following the conclusion of the third conference call,
which I understand to have taken place on the morning of Wednesday
31st July, it became unclear as to whether and to what extent material
was either being retained centrally or disseminated. The following
information has been relayed to me:
i, The minutes of a previous conference call had been typed and
emailed to a number of persons. An instruction was then given
that those emails and minutes should be, and have been,
destroyed: the word “shredded” was conveyed to me.
ji. Handwritten minutes were not to be typed and should be
forwarded to POL Head of Security.
iii. Advice had been given to POL which I report as relayed to me
verbatim: “If it's not minuted it’s not in the public domain and
therefore not disclosable.” “If it’s produced its available for
disclosure — if not minuted then technically its not.”
iv. Some at POL do not wish to minute the weekly conference
calls.”
115. I note that these events appear to have followed the third Horizon Regular Call
on 31 July 2013, which I did not attend,®* and I was not aware of these events
prior to receipt of the Second Clarke Advice
87 POL00229411.
88 POL00193767.
Page 67 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
116. The Second Clarke Advice warned that if these allegations were well-founded
and potentially disclosable information or material had been lost or destroyed as
a result, this would amount to a serious breach of POL’s duties as prosecutor,
and (depending on the circumstances) could amount to conspiracy to pervert the
course of justice by those involved. It concluded that the “proper way forward is
for the conference calls to be properly minuted, those minutes to be centrally
retained and made available to all those who properly require access thereto”.
Alternatively, it suggested that “some other centrally-based mechanism be
designed, so as to permit the collation of all Horizon-related defects, bugs,
complaints, queries and Fujitsu remedies, arising from all sources, into one
location. Such a mechanism would amount to proper compliance with that aspect
of a prosecutor’s duty relating to the recording and retention of relevant
information’. \t was this latter suggestion that Susan Crichton and Hugh
Flemington called me to discuss.
117. On reviewing the Second Clarke Advice, I thought that the matters it dealt with
were serious, and I understood the central point of Cartwright King’s advice.
However, as Susan Crichton and Hugh Flemington had observed, Cartwright
King’s advice did ask a lot of POL. It asked POL not only to record bugs, but also
“complaints and queries’, presumably even those that did not turn out to reveal
bugs. It also required this to be captured “from all sources” which I thought was
a substantial undertaking given that POL had several thousand SPMs who could
contact POL through a variety of channels, let alone its customers, clients and
its own staff who may also raise Horizon-related “queries”). I sent, in summary
form, these thoughts to Susan Crichton:
“The bit of the advice that concerns me is:
Page 68 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
“I would advise that either the conference calls be continued or
that some other centrally based mechanism be designed, so as
to permit the collation of all Horizon-related defects, bugs,
complaints, queries and Fujitsu remedies, arising from_all
sources, into one location. Such a mechanism would amount to
proper compliance with that aspect of a prosecutor's duty relating
to the recording and retention of relevant information” (emphasis
as in my original).
This approach is very robust but the question is whether this is workable
in practice? Perhaps CK could be asked to consider if there is an easier
but still defensible way to meet the disclosure duty?”
118. In the event the Horizon Regular Calls continued albeit with a better-defined
system for recording information in place, which included WBD taking on the role
of keeping minutes. The way in which this was established was as follows:
118.1.Following the above exchange with Susan Crichton, I agreed with Dave
Posnett of POL that the appropriate way of keeping the minutes for the
calls would be to add them to a spreadsheet that I (or WBD) was to
prepare. This single spreadsheet would then comprise a complete record
of all calls, and each week an updated version would be circulated to
attendees before the next call. This agreement is reflected in Dave
Posnett’s email to me of 16 August 2013, which is shown in
POL00139691. I cannot specifically recall the conversation referred to, but
the process embodied in the email is the one I remember.®°
118.2.Dave Posnett forwarded the notes that he and his Security Team
colleagues at POL had made of the calls to date, in order for me to place
®9 For completeness, the email from Rob King to Jarnail Singh shown at the top of POL00139691
is not one I recognise and it has not been possible to locate it on my firm’s file; it appears to
have been internal to POL.
Page 69 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
them within the first iteration of the spreadsheet. This was the first time I
had received notes of these earlier instances of the Horizon Regular Call.
118.3.With reference to Q25.1 of the Request, I received an email from Jarnail
Singh on 20 August 2013 (POL00139693) about how information and
action points from the calls was being retained. I explained that WBD was
collating all the minutes prepared by POL into a single "Weekly Report"
(which was in fact a Word document rather than a spreadsheet). I also
explained that the action points from each meeting were recorded in the
minutes. Still, I recall being rather perplexed by Jarnail Singh’s email. Its
phrasing was unclear, and it seemed to be seeking substantive information
even though this was a process initiated by POL and Cartwright King, and
WBD had just been asked to take over essentially a secretarial function of
capturing the minutes into a single place.
118.4.1 then emailed Jon Scott to seek his agreement to me circulating the
Weekly Report to the attendees of the call later that day. He responded
agreeing to that and saying that he “would suggest that all the
issues/matters raised are also collated onto one action sheet’.
118.5.1 circulated the first Weekly Report, incorporating all of the minutes from
the first five Horizon Regular Calls, on 21 August 2013 shortly before that
day's call.91
118.6.With effect from the 21 August call, WBD provided a paralegal to take the
minutes. My weak recollection is that shortly before this point someone
°° See for example POL00137427.
®1 WBONO000796.
Page 70 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
from POL legal asked me if WBD could provide a paralegal to keep
minutes of the calls going forward and that we were asked to do this as
we had the resources to ensure this was done routinely every week. So
at this stage, WBD's role was to take the minutes of each call and then
collate them into the Weekly Report.
118.7.On 22 August 2013, Hugh Flemington of POL asked me and Rob King of
POL who was doing an "action log". I agreed that the WBD paralegal doing
the minutes could also take charge of maintaining an action log to be
updated each week, subject to receiving clear directions from POL as to
what actions were to be recorded and how they were to be described.°? I
am asked by Q26.2 of the Request to describe a telephone conversation
I had with Jarnail Singh and the Security Team regarding the “action point
list’. I have no independent recollection of this, but the above exchange
dated 22 August 2013 suggests that its purpose was to agree a format for
the action log in line with John Scott's email on 21 August 2013 that there
should be one "action point sheet".
Protocol
119. By Q27 I am asked to consider a “Protocol” (POL000139696) circulated by email
to Horizon Regular Call attendees (including me) on 9 October 2013
(POL000139695). I recognise this as a document drafted by Cartwright King but
otherwise have no specific recollection of it.
120. Having reviewed it for the purposes preparing this statement, I note that it broadly
summarises my understanding, explained above, as to why the Horizon Regular
°2 WBON0000807.
Page 71 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Call was set up (namely, to aid POL to comply with its prosecutorial disclosure
duties) and how it was to be managed. I do not know why it included a
requirement for a solicitor representing WBD to attend “each and every” Horizon
Regular Call, along with a note-taker (§3.3), or why it specifically required the
retention of the minutes of the Horizon Regular Call for a period of 6 years
(§4.5.3) — the latter was something that WBD was doing in any event in line with
its usual file retention policies. Also a WBD solicitor generally attended (at this
stage, often me), it being prudent to make sure that we were sighted on the calls
and could feed in relevant observations from the Mediation Scheme workstream.
As explained above, from mid-August 2013 a WBD paralegal took on the role of
taking minutes and maintaining an action log and we would naturally retain
documents arising out of the calls as part of our ordinary document retention
policies. I do not think that the points in the Protocol were specifically agreed with
WBD, but for the reasons given above I doubt we would have particularly
objected to them if asked.
121. As to why the Protocol was created more generally, again I have limited
independent recollection of this. As set out above, by early October when this
document was circulated the elements principally affecting WBD -— that is, the
procedure for recording minutes and actions — had been settled. I do note that in
WBD’s note of the 9 September 2013 conference with Brian Altman QC
(POL00006485), discussed further below at §§ 123 ff, Simon Clarke of Cartwright
King is recorded as saying that “he thought that it was necessary to put duties
on individuals. Consequently CK are in the process of writing a protocol to explain
the purpose of the weekly hub meetings, the roles and responsibilities of
individuals”. I have no reason to doubt that this is a fair reflection of his motivation
Page 72 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
in preparing the Protocol. I note also that this comment was in the context of him
saying that there were some “cultural issues” at the outset of the Horizon Regular
Calls. I do not recall this comment but it may have been a reference to the
allegations cited in the Second Clarke Advice, of which I was not aware until
reading that advice. It may also, or alternatively, have been a reference to the
fact that there were some difficulties in establishing a process for the Horizon
Regular Call and progressing actions. To the best of my recollection these were
ordinary and minor teething issues, i.e. it took a few calls to get into a proper
routine as to who would regularly attend on behalf of the different departments
involved, and there was not always clarity as to who would be responsible for
progressing actions arising out of the meetings. I would also note that at this
early stage, WBD and I had very limited visibility of the prosecution side of POL,
the security team and Cartwright King's work so there may have been other
issues in play that I was not sighted on.
122. Fairly early on in the life of the Horizon Regular Calls — which to my knowledge
continued for several years — I stopped attending. As I have set out above, my
firm’s file indicates that the last call I attended was on 19 February 2014, with
Claire Parmenter (a solicitor at WBD) attending some of the Horizon Regular
Calls after I had stopped. By June 2014, Claire had left the firm and thereafter
different people will have been the usual WBD attendee at different points in time.
For example, I am aware from my firm’s file that after Clare left her successor on
the calls was Alva Leigh-Doyle. At some point thereafter I recall that WBD’s
attendance on the calls dropped from a solicitor attending to a paralegal, though
I cannot remember when this happened.
Page 73 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(v) Conference with Brian Altman QC on 9 September 2013 (Q30)
123. The prelude to the conference on 9 September 2013 was the delivery of Brian
Altman QC’s interim advice dated 2 August 2013 (the “BAQC Interim Advice”).°°
This was a preliminary report on the propriety of the parameters that Cartwright
King had set in undertaking their review of historic convictions, and on the
general approach Cartwright King were taking in relation to that review. It was
provided to Gavin Matthews and Simon Richardson (WBD’s client relationship
Partner for POL), who then forwarded it to me on 4 August 2013.
124. In short, Brian Altman QC did not raise fundamental concerns about the ambit of
or approach to the review, but he did identify certain areas where further thought
might be required. For example: he queried whether the review should go back
further than three years, Cartwright King having identified 2010 as the cut-off
date for convictions which were to be the subject of their review (§ 15); he thought
that consideration may need to be given to whether there were other issues,
beyond Gareth Jenkins’ non-disclosure, which could potentially give rise to
grounds for appeal in cases subject to the review (particularly in light of the issues
identified in Second Sight’s Interim Report) (§24); and he gave some constructive
feedback on Cartwright King’s approach and what could be done to avoid
potential pitfalls (§24).
125. In view of these points, Simon Richardson provided some brief advice to POL as
to what its next steps should be. In summary he said:
°° POL00223376.
°4 WBON0000393.
Page 74 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
“1.POL legal needs to disclose Brian Altman's Interim Review to CK and
discuss it with them.
2. CK should be asked to respond in writing to the recommendations
made at Paragraph 24 and Paragraph 15
3. Bond Dickinson (Andy Parsons) should sit down with Brian Altman to
walk him through the spot review process and the SS Report so that he
can understand the impact of his review on the civil side”.°>
126. Of relevance for my purposes, Simon also observed that Brian Altman QC “raises
the issue of whether the current review is too narrow ... he references the list of
issues in the SS report and Spot Review 22 as examples of other issues which
may need to fall within the ambit of CK's review’. He commented that “/wJhilst
this should be put to CK, my own view is that it may be very difficult for CK to
expand the review on issues on which SS have failed to come to any conclusion.”
127. From this I understood that Simon was concerned to ensure that Brian Altman
QC had a full understanding of the Spot Review process and the challenges
posed by the fact that Second Sight had only made limited progress in
investigating the issues raised by these cases. Given that the Second Sight
investigation was potentially material to his views on how the review should be
conducted, and given that the investigation was in a complex transitional phase
as the details of the Mediation Scheme were being worked out, it made sense
for me to sit down with Brian Altman QC to talk him through this workstream (on
which I was leading), notwithstanding that I was not especially close to his
instruction.
°5 WBONO000786.
Page 75 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
128. I was also aware there was a degree of overlap between the civil and criminal
workstreams at this point, as a number of the cases that were under
consideration by Second Sight involved SPMs who had also been convicted. I
was therefore concerned to understand what was happening in the criminal
workstream so as to identify any possible impacts on the civil side (for example,
civil claims that might flow from SPMs’ convictions being found to be unsafe).
129. In addition to this, there was a need for Brian Altman QC to meet with Cartwright
King around this time, in order for them to discuss the points arising out of the
BAQC Interim Advice with him. From my emails, I can see that they prepared a
written response to the Interim Advice, which was forwarded to me by Susan
Crichton on 13 August 2013% (together with the Second Clarke Advice,
discussed above at §§112-118). In line with the rationale for WBD rather than
Cartwright King acting as Brian Altman QC’s instructing solicitors for the purpose
of carrying out his review, it was generally felt that WBD should be present at any
meeting between the two. A conference resulted at which members of WBD,
Cartwright King and POL’s legal team were all present. This was the conference
of 9 September 2013 (the “Conference’”).
130. I have been asked to set out on my recollection of this conference. Other than
the fact that I helped to arrange it, that it took place at Mr Altman KC’s chambers,
and the broad reasons for my attendance as set out above, I cannot recall the
specifics of what was discussed. I can therefore only comment by reference to
the two notes of it which the Inquiry has provided — POL00006485 and
POL00139866.
86 WBON0000791; POL00223376.
Page 76 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
131. The first of those notes (the “WBD Note”) is headed with the name of the firm,
and the metadata that my firm has managed to obtain from our electronic filing
system suggests that it was typed up by a secretary on Gavin's behalf. My firm's
records show that Gavin circulated the WBD Note with me in copy following the
9 September Conference on 23 September 2013.” The “Second Note” does not
appear to me to be a WBD document. That assessment is supported by the
absence of the Second Note in WBD’s records. Therefore, I have no reason to
believe that I have seen the Second Note prior to preparing this statement and I
do not know who prepared it. I do not specifically recall what was discussed at
the Conference but I have no reason to doubt the accuracy of either Note, which
appear broadly consistent in terms of the topics discussed and order of play.
132. I am asked, first, to comment on the references to ‘cultural’ issues at the outset
of the Horizon Regular Calls. I have dealt with this above at §121 in the context
of the Protocol which was later drafted by Cartwright King to govern those calls.
I am unable to say what was meant by the comment in POL00139866 attributed
to Susan Crichton (“People then dump...”), which occurs at this part of the
Conference.
133. Second, I am asked to set out my recollection of the part of the discussion
relating to Gareth Jenkins. I cannot recall this discussion and can only comment
that, as I have already set out, Cartwright King had by this time come to the view
that Gareth Jenkins could not be relied upon as an expert witness as he had
failed to disclose material information about problems in the operation of Horizon
Online when giving evidence in previous criminal prosecutions. I recall that Mr
87 POL00333840.
Page 77 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Altman QC had reached a similar view in his (then) recent Interim Advice, so this
part of the conference was probably about that.
134. Third, I am asked to expand on Brian Altman QC’s advice, recorded in
POL00006485, that POL had “no positive duty to seek out individuals pre 1
January 2010 but if [it] was approached it would need to make case-specific
decisions on disclosure”. I recall in general terms that Brian Altman QC approved
Cartwright King’s decision not to proactively review cases where the SPM’s
conviction had been imposed prior to 2010 (having previously raised this as a
discussion point in his Interim Advice). However, I cannot recall the specific
rationale for his view that it was sufficient for POL not to proactively investigate
this category of case to see if there had been a failure of disclosure. I do not
know the source for the statement in the WBD Note that “/pJrior to the HOL rollout
there was a cash audit done so that all POL branches balanced”; and I do not
understand how this bore on Brian Altman QC’s reasoning.
135. Fourth, as to whether the Callendar Square bug was discussed in this context, I
cannot see from the two notes of the Conference that it was (nor can I recall this).
However, in the course of preparing this statement I have located a handwritten
note (written by me) of a conference which appears to have taken place on 4
October 2013 with Brian Altman QC, me, Gavin, Cartwright King and members
of the POL legal team present. It is a short note and the conference it records
appears to be in the nature of a brief follow-up call to discuss a discrete point.
That point was (it seems) whether 12 cases involving convicted SPMs who had
applied to the Mediation Scheme should be reviewed by Cartwright King, i.e.
notwithstanding that they were outside the current scope of that review. I have
no memory of this conference and was likely there because it concerned a
Page 78 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
specific issue about Mediation Scheme applicants (it not being any part my role
to advise on the proper scope of Cartwright King’s review). I can therefore do no
more than draw the Inquiry’s attention to what the note records:
“GM: ... These cases were not in CK full review.
Q: Should they be reviewed?
Current answer: No cases before 1 Jan 2010 (Horizon Online).
Also, Falkirk bug which was before 1 Jan 2010.
MS: Q is whether Falkirk bug affected other cases and whether further
disc needed?
BAQC: Letter G p46 GJ says the problem was at Callendar Square.
Affected in 2005.
Fix rolled out in March 2006 — network wide.
1 Jan 2010 — logical and proportionate.
If D’s say prob with Old H, then can review cases on an ad hoc basis.
SS were not limited to HOL or Old Horizon.
And no bugs found by SS in Old H.
So on solid ground to stop review at 1 Jan 2010.”
136. Fifth, I do not specifically recall Brian Altman QC’s statement, recorded in
POL00006485, which identified the concern that “lawyers acting for [convicted
individuals] may be using the [mediation] scheme to obtain information which
they would not normally be entitled to in order to pursue an appeal’. Having
reviewed this record in context, I surmise that his apprehensiveness centred on
SPMs getting hold of information through the Mediation Scheme before the
criminal legal team had an opportunity to review and formally disclose it through
88 WBONO000725.
Page 79 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
the conventional prosecution channels. More generally (though I cannot recall to
what extent this point was discussed at the 9 September Conference), I recall
that both Cartwright King and Mr Altman QC were of the view that the Mediation
Scheme should not be open to SPMs whose convictions had not been quashed,
as they thought that POL mediating these cases sat uncomfortably with the fact
that the status quo was that these convictions were sound.
(v) Susan Crichton’s departure (Q33)
137. Lastly in this section, I deal with the Inquiry's Q33 (which asks what I thought the
reasons were for Susan Crichton’s departure from POL). I do so for convenience
for the simple reason that this happened at around this point in time (i.e. in or
around September 2013). In short, I was not aware of the reasons for this and
still am not. I would not expect to have been given any details (and I do not
believe I would have asked), as I had only been working closely with her for a
few weeks and was still comparatively junior. I simply knew that Susan Crichton
had left and that later Chris Aujard was appointed to replace her as POL’s interim
General Counsel.
G. MEDIATION SCHEME (Q32, Q35 to Q52)
138. This section addresses the Mediation Scheme, the background to the
establishment of which I have explained above. When considering the Inquiry’s
questions about the scheme (which are, broadly speaking, Q35 to Q52 of the
Page 80 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Request),%? I have found it helpful to think of matters in terms of six broad topics.
These topics do not always follow the order in which the Inquiry’s questions are
set out, so for convenience I set out the structure I have adopted here:
138.1.In subsection (i), I give an overview of the Mediation Scheme and the
nature of the work my firm and I were instructed to carry out in relation to
it. This answers Q35, as well as Q36 and (in part) Q38.
138.2.Next, the Inquiry has asked various questions about the process that was
followed by POL during investigations into complaints submitted by SPMs
— in particular, the Inquiry has queried POL’s reasons for issuing a civil
claim against an SPM whilst the investigation into his case was ongoing —
and I am asked about concerns which were raised during the early stages
of the scheme in relation to the timeliness and quality of POL’s POIRs.
These questions are Q37 and Q39 to Q40, and Q42.1, which I answer in
subsection (ii) below.
138.3.In subsection (iii), I answer various questions raised by the Inquiry in Q41
and Q42.2 to Q44 of the Request, concerning my views on the work
carried out by Second Sight during the Mediation Scheme.
138.4.The Inquiry has also asked about the provision of information to SPMs and
Second Sight during the Mediation Scheme, and specifically, POL’s
approach to providing: the Helen Rose Report; ‘Officer's Reports’; and
information about a form of remote access known as the ‘Balancing
°9 Additionally, Q34 refers me to a series of minutes and action logs arising out of meetings of the
Working Group, which I have reviewed; and I answer Q32 of the Request in this section as
explained further below.
Page 81 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Transaction’ functionality (Q32 and Q49, Q46, and Q45 of the Request,
respectively). I answer these points in subsections (iv)-(vi) below.
138.5.In subsection (vii), I set out my (and WBD’s) role in advising POL on the
merits of applicants’ cases and whether or not to take a case to mediation.
This answers Q38, as well as Q47 to Q48.
138.6.Finally, I address POL’s decision to close the Working Group in March
2015 (Q35.6 and Q50 to Q52).
(i) Overview of the Mediation Scheme and my / WBD’s role in relation to it (Q35
to Q36, Q38)
The Mediation Scheme and Working Group
139. As set out above, the Mediation Scheme opened on 27 August 2013 and closed
to new applications on 18 November 2013. I recall that approximately 150
applications were received in that window, which was more than POL had been
anticipating — although a handful of these were not ultimately accepted onto the
scheme by the Working Group.
140. The scheme had a two-part structure (investigation followed in some cases by
mediation), which reflected that its dual purpose was (i) to offer a mechanism for
investigating eligible complaints by SPMs (which was hoped sufficient of itself to
dispose of some complaints by giving the SPM greater insight into POL’s
decision-making in their case), and (ii) to provide for the mediation of cases
deemed suitable following this initial investigation. In line with this structure, the
essential steps in the scheme were as follows:
Page 82 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
140.1.The SPM would apply to the scheme and the Working Group (whose role
I describe further below) would decide whether or not they should be
accepted onto the scheme for further investigation in line with the
scheme's eligibility criteria.
140.2.Second Sight would send the SPM a case questionnaire.
140.3.The SPM prepared a Case Questionnaire Response (“CQR”). POL would
pay for a professional advisor, usually a lawyer or accountant, to assist in
the preparation of the CQR. This was because sometimes SPMs’
concerns and criticisms were not articulated clearly, which was
understandable given the complexity and length of time that had passed
in some cases. POL believed that providing SPMs with professional
advice would help them explain their concerns, which in turn would help
Second Sight and POL investigate them.
140.4.The CQR was returned to Second Sight. POL did not have full visibility of
this part of the process, but I was aware that Second Sight would
sometimes send the CQR back to the SPM asking them to provide more
information. In due course it appeared to me that Second Sight was getting
increasingly involved in helping SPMs to draft their CQRs (something they
called "hardening" the CQRs), and I deal with the upshot of this further
below.
140.5.The CQR was sent to POL to investigate and prepare a response in the
form of a POL Investigation Report (“POIR”). Although there was no formal
or legal requirement for disclosure in the context of the Mediation Scheme
(as it was a voluntary process and not a form of civil proceedings), POL
Page 83 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
provided a pack of relevant evidence from its own files together with the
POIR. POL put together a team of investigators for this purpose, who
investigated each issue raised by the CQR, collated relevant evidence,
and produced the draft POIR. The in-house investigative team was led by
Kathryn Alexander and Shirley Hailstones under the management of
Angela Van Den Bogerd. Kathryn Alexander and Shirley Hailstones had
deep experience of working in branches, and Angela Van Den Bogerd also
had very detailed knowledge of how SPMs operated and was thought by
POL to be someone who could communicate well with SPMs).1°
140.6.The POIR would be passed to Second Sight by POL, who were to review
the input from both the SPM and POL and produce their own report as to
the merits of the case and whether it should proceed to mediation.
140.7.Second Sight prepared a Case Review Report (“CRR’) in draft form which
was sent to POL and the SPM for comment. The comments were then
reviewed by Second Sight and a final report was produced.
140.8.The report was reviewed by the Working Group which considered whether
a case should be recommended for mediation.
140.9. Where it was agreed that a case would proceed to mediation, a case file
was prepared and sent to the Centre for Effective Dispute Resolution
("CEDR’) which was engaged to conduct the mediations. Mediations took
10° In Second Sight's Part 2 Briefing Report they recorded that "we wish to place on record our
appreciation for the hard work and professionalism of Post Office’s in-house team of
investigators, working for Angela Van Den Bogerd, Post Office’s Head of Partnerships. Our
work would have been much harder and taken much longer without the high quality work carried
out by this team. We have also received excellent support from the administrative team set up
by Post Office to support the Working Group” (paragraph 26.5).
Page 84 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
place face-to-face between POL and SPMs, and the SPM was entitled to
be accompanied by their professional advisor who was funded by POL.
141. The Working Group’s role was to consider whether a case was suitable to
proceed to mediation, taking into account (though it was not bound by) Second
Sight’s recommendation in its CRR. As I explain further below, during the course
of the scheme it was decided that in cases where POL and JFSA did not agree
on the suitability of a case for mediation, the Working Group’s consideration of
this issue would be formalised in a vote. The Working Group's voting structure
(for this and other decisions that fell to the group) was that POL and the JFSA
each had one vote and Sir Anthony Hooper had the casting vote in the event of
a tie. It was not part of the Working Group's function, however, to render any
opinion on the substantive merits of SPMs’ cases.
142. More broadly, the Working Group’s role was to oversee the administration and
operation of the Mediation Scheme, and in particular to: (i) ensure the timely
progression of SPMs’ complaints through the investigative phase of the scheme,
including deciding requests for extensions of time by participants to prepare
CQRs, POIRs and CRRs (as applicable); (ii) consider requests from SPMs for
extra financial assistance (i.e. over and above the baseline level provided by
POL), which POL would meet if approved; and (iii) deal with any other ‘process’
issues which arose. The workings of the Working Group and the content of its
discussions were subject to confidentiality and without prejudice privilege.
143. Second Sight’s role within this process was to investigate SPMs’ complaints
independently and impartially, reviewing the information and evidence provided
by both sides (and pursuing follow-up enquiries if necessary) in order to give a
logical and fully evidenced opinion on the merits of the SPM's complaint and a
Page 85 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
view on whether the case was suitable for mediation. If it went to mediation,
Second Sight’s case-specific findings would form part of the case file for use by
the parties and mediator. Second Sight were also commissioned by the Working
Group to prepare two general (as opposed to case-specific) reports: the first of
which was to be a neutral, objective overview of the key elements of the Horizon
system and associated processes, including training and support processes (this
came to be known as the “Part 1 Briefing Report”); and the second was to be a
‘thematic’ report dealing with commonalities that Second Sight had identified
between different SPMs’ cases in the course of their investigations (the “Part 2
Briefing Report”). Both reports were conceived as briefing reports for use by the
mediators to help them understand the background context to individual cases,
and in that sense differed from the report which Second Sight had originally been
commissioned in 2012 to produce to identify whether there was a system-wide
problem with the Horizon software.
144. I recall that it was originally envisaged that Second Sight would also handle the
administrative aspects of the Mediation Scheme, e.g. writing letters to SPMs to
notify them of deadlines and case updates. However, given that a larger than
expected number of applications were received during the application window, it
was decided that the administrative aspects of the scheme should be run by POL.
POL in turn brought in a team of external consultants to help administer the
scheme alongside their own staff. The Mediation Scheme came to be known as
Project Sparrow within POL and POL’s participation in the scheme was managed
by the newly created in-house team at POL, which Belinda Crowe was appointed
to lead. This was the team (alongside POL's in-house legal team) that I interacted
with and took instructions from. I had limited visibility of the governance structure
Page 86 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
and decision making above this team. I do not recall attending any Board or
Board Subcommittee meetings about the scheme.
145. The Working Group met formally roughly every four to six weeks, with additional
shorter conferences (usually by telephone) in between. It consisted of Sir
Anthony Hooper, POL and JFSA (who could bring multiple attendees, though
organisationally only had one vote), and Second Sight (who sat in a non-voting
capacity). The Working Group’s regular attendees were: Sir Anthony Hooper;
Alan Bates and Kay Linnell from the JFSA; Ron Warmington and lan Henderson
(and later, also Chris Holyoak) from Second Sight; POL's General Counsel;
Belinda Crowe and Angela Van Den Bogerd; and me. Others from POL
occasionally attended too. Susan Crichton attended the first few meetings
between the above individuals as General Counsel, but I cannot recall if these
were formally constituted as the Working Group at that point. When she left, Chris
Aujard took over as General Counsel and then at the beginning of 2015 he was
replaced by Jane MacLeod.
146. The Working Group oversaw the Scheme until March 2015, when POL decided
to disband the group and mediate all cases within the scheme which had not yet
been the subject of a decision by the Working Group, save for those where the
SPM had an extant criminal conviction. I deal with this decision in more detail
below. Thereafter there was a run-off period during which Second Sight
continued to produce CRRs in relation to individual cases and the outstanding
cases were mediated. The records held on my firm’s file suggest that the last
mediation in which WBD was directly involved took place in 2016 (although my
involvement was largely complete by mid-2015 or thereabouts).
Page 87 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
147. It should be noted that the above brief narrative is provided by way of overview
only, so to contextualise my explanation below of the nature of my (and my firm’s)
instructions in relation to the Mediation Scheme. It by no means provides a
comprehensive account of how the scheme and Working Group functioned at
every stage throughout their lifetime, which I address (so far as relevant to the
Inquiry’s questions) in the sections that follow.
My / WBD’s role
148. When the Mediation Scheme was running, I was a Senior Associate. I had day-
to-day conduct of the Mediation Scheme instruction on behalf of WBD, working
under the supervision of Gavin Matthews.
149. My role (and where applicable that of the wider firm) was as follows:
149.1.1 attended Working Group meetings on POL's behalf. This meant that, on
occasion, I would present POL’s position on a particular agenda item.
From time to time I would provide assistance to the Working Group itself,
for example, drafting letters to be sent out by the Working Group or in Sir
Anthony’s name, or redrafting aspects of the Working Group’s terms of
reference.
149.2.1 managed the team of WBD lawyers and paralegals who were supporting
POL’s work investigating and advising on complaints that were submitted
to the Mediation Scheme. The team’s work broadly comprised the
following:
(i) They reviewed each CQR and for most of them prepared a list of issues
to help guide POL’s investigations and they reviewed draft POIRs and
redrafted them where necessary to ready them for submission to
Page 88 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Second Sight. The WBD team’s work did not include carrying out any
investigative work in relation to SPMs’ cases. This was done by POL’s
in-house investigative team. WBD’s role was focused on ensuring that
the POIR was clearly written, did not contain any obvious errors, and
addressed each of the issues raised by the CQR. As set out further
below, early on there were some concerns about the quality of POL’s
POIRs and at that stage WBD dedicated extra time to reviewing draft
POIRs (in particular, by preparing executive summaries and reviewing
the underlying evidence relied on by the POL investigative team to
ensure that the conclusions in the POIR were more clearly and closely
tied to the evidential output of POL's investigation). I personally
reviewed most (if not all) of the POIRs prior to submission to Second
Sight. My recollection is that all the POIRs were sent to Cartwright King
for review. Final review and sign-off of all POIRs was done by Angela
Van Den Bogerd and POL’s in-house legal team. Initially this fell to
Rodric Williams but this led to a backlog due to constraints on his
capacity, so in due course Jonny Gribben (a WBD solicitor who was
then on secondment at POL) took on this role.
(ii) The WBD team advised on the merits of each claim once Second Sight
had produced its CRR. That is, the team would provide concise written
advice on whether POL should agree to mediate the claim, and if so,
what the settlement parameters were. These advice notes were based
on, and applied: (i) advice given to POL by Linklaters on 20 March 2014
to the effect that, absent proof that Horizon was not working as it
should, POL was contractually entitled to recover losses which the
Page 89 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Horizon system recorded as due and owing (I address this advice
further below at §248); (ii) advice given by POL’s criminal lawyers that
it should not mediate with any SPM who had an extant criminal
conviction (I address this further below at §§246-247); and (iii) a
document setting out POL’s general settlement criteria and approach
to valuing claims which I prepared. I believe I reviewed each advice
note produced by the team before it was provided to POL.1°'
(iii) If a case went to mediation, a lawyer from WBD would attend in person
to represent POL. I recall attending two or three mediations personally,
but generally this was done by the members of the WBD team whom I
supervised.
149.3.The WBD team also assisted POL in preparing written material of a
generic (as opposed to case-specific) nature to assist Second Sight in
carrying out its work. As I recall this had three main components: first,
WBD helped to draft the “Horizon Factfile” document (which I come back
to below); second, we helped to draft notes about particular aspects of the
Horizon system, associated processes, and POL’s business practices, in
response to queries from Second Sight and based on information provided
by POL and/or Fujitsu;'°? and third, WBD helped POL to prepare a long-
form paper responding to questions posed by Second Sight about the
101 WBONO001702
102 See POL00201950. I refer to an example of such an advice note (about the ‘Balancing
Transactions’ functionality) below at §§208 ff, albeit that this particular note was never ultimately
finalised.
Page 90 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
‘thematic’ issues it had identified in the course of its investigations, to
assist it in finalising its Part 2 Briefing Report.1°
149.4.From time to time, when requested by the POL legal team, I also fed in my
views on wider issues. For example, I reviewed and commented on
Second Sight’s generic or ‘thematic’ reports. Towards the end of the
scheme, by which time WBD had assumed a greater role, I commented
on a paper on the closure of the Working Group that was to be put before
the Project Sparrow Subcommittee (see below, §§255-259). As mentioned
above (§149.2(ii)), I also produced, with WBD colleagues, a document
setting out POL’s general criteria for settling SPM’s claims and guidance
on quantum. This approach was broadly adopted by POL, subject to and
in light of the advice it received from Linklaters and its criminal law advisers
to which I have referred above.
150. I was formally instructed by POL's General Counsel, which during this period was
mainly Chris Aujard. However, on a day-to-day basis my instructions would
typically come from Rodric Williams, Belinda Crowe and Angela Van Den Bogerd
in relation to matters which were within their spheres of responsibility.
The Horizon Factfile
151. Q36 asks me about my/WBD’s role in the Horizon Factfile document. To the best
of my recollection, the genesis of the Horizon Factfile was that the Working Group
agreed that POL would provide a neutral, objective document describing the key
features of the Horizon system and its associated processes, in order to (i) help
cut down the amount of drafting time required to prepare reports on individual
103 Appended to Second Sight's Part 2 Briefing Report.
Page 91 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
cases, and (ii) assist Second Sight in preparing its Part 1 Briefing Report.
However, I cannot remember the exact sequence of events and it may be that
POL was already preparing such a document for its internal use when it offered
it to Second Sight to draw on in preparing its Part 1 report.1™
152. As I have mentioned above, WBD assisted POL in drafting this document. Claire
Parmenter (then a solicitor at WBD) took day-to-day responsibility for this work
under my supervision, collating information from various departments and teams
within POL to include in the Factfile.1°° Needless to say, the information contained
within the Horizon Factfile was included on instruction given by POL to WBD.
Having reviewed Rodric Williams’ email of 20 December 2013 to which I am
referred by Q36 (POL00021860), I believe that this email was intended to outline
the factual areas Rodric Williams initially identified as needing to go into the
Factfile, and to name the individuals within POL who would be likely to be best
suited to contribute the relevant information. I note that one of the areas identified
was “Branch Settlement’ (and within that, “how resolved”), which was allocated
to Rod Ismay (Head of Finance Service Centre, POL). I further note that in the
draft of the Factfile at POL00040066, the title of the section in which §41.3
appears refers to Rod Ismay. I would therefore presume that Rod Ismay (or his
team) supplied the information contained in §41.3 of that draft, although I cannot
say for certain whether he, someone else at POL, or a WBD lawyer held the pen
on that specific paragraph.
104 Cf, POL00021860 and POL00026656. See also WBON0000824 (minutes of Working Group
meeting on 1 April 2014 where handover of the Horizon Factfile to Second Sight was
discussed).
105 POL00021860; see also this email from Claire to me dated 10 January 2013, which outlines
the general approach she took to compiling the 8 January 2013 draft (i.e. POL00040066):
WBONO0000396.
Page 92 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
153. This inference appears to be supported by an email chain I have identified from
my firm’s file, which Claire Parmenter sent to me on 11 February 2014.1 That
chain shows that on 30 January 2014, Claire had sent Rod Ismay a draft Factfile,
asking him to “review the section [he] helped to complete” and to “confirm
whether [he is] happy with the wording (and let [her] have any amendments)”. I
have compared the draft Factfile Claire to Rod Ismay for approval and the content
of the relevant paragraph is largely identical to that in POL0004066.1°” She
recorded that the section Rod Ismay had assisted with was “Branch Settlement
- pages 12-15” 108
154. For completeness, I note from this chain that her email was then circulated within
Rod Ismay’s team and Andy Winn responded with the following comment:
“41.4 Settle centrally (>£150) and dispute the shortage - if the
subpostmaster believes that the shortage was not his/her fault or could
be resolved through other means (see below), then the debt will be
suspended to allow time for the shortage to investigated and remedied.
The subpostmaster disputes a shortage by contacting the Network
Business Service centre (NBSC), Cash Centre (remittance disputes) or
Finance Service Centre ("FSC") for transaction corrections at Post
Office.”
155. Andy Winn’s comments were then forwarded to Claire by Rod Ismay on 11
February 2014. I have identified that, subsequently, an updated draft of the
Horizon Factfile was circulated by me to Belinda Crowe and another on 21
February 2014.19 The version of the relevant paragraph which appears in that
draft (at §47.3) is in substantially the same terms as Andy Winn’s text quoted
16 WBONO000398.
107 See: WBON0000401. The main difference appears to be that the comments are not on
POL0004066.
108 WWBONO000402.
108 WBON0000812.
Page 93 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
above.""® In a later email in the same chain, Claire confirmed that Rod Ismay had
reviewed the relevant text, i.e. the text contained in the section titled “Branch
Reporting & Management (pages 13 to 17)”.11"
(ii) Process issues during POL’s investigations into complaints (Q37, Q39 to
Q40, Q42.1)
Proceedings issued against Terence Walters
156. POL00026666 (an actions list arising from the Working Group meeting on 12
December 2013) refers to a claim issued by POL against Terence Walters, who
was applicant M006 in the Mediation Scheme. In answer to Q37 of the Request,
these proceedings were issued on 28 November 2013 in order to protect POL’s
limitation position, as limitation was due to expire the following day.''? A letter
was sent to Mr Walters explaining this to him and offering to immediately stay the
proceedings in light of his application to the Mediation Scheme, “so to assure
[him] that no further action will be taken at this time”.""> I emailed the Working
Group on 13 December to explain the steps POL had taken and that it proposed
to stay the proceedings."
157. The claim was stayed by consent for 6 months on 29 January 2014 to allow the
mediation process to complete,''® and on 9 September 2014 the stay was further
extended.'® Ultimately the claim remained stayed until it was dismissed by
110 WBONO000813.
111 WBONO000814.
112 WBONO000808. The Notice of Issue is: WBON0000951.
118 WBONO0000809.
114 WBONO0001670.
‘18 Consent Order: WBON0000950.
116 See WBON0000890; the Consent Order is: WBON0000949.
Page 94 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
consent on 13 October 2020, Mr Walters having been one of the claimants in the
group litigation.11”
Issues with POL’s early POIRs
158. Q39 of the Request asks me to set out my recollection of the discussion at the
Working Group meeting on 13 March 2014 as to the number of extensions of
time that POL was seeking to prepare its POIRs (POL00026643). I have no
specific memory of this meeting or of the discussion triggered by Alan Bates’
remarks, but I do recall that around the beginning of 2014 there were generally
delays in progressing SPMs’ cases through the scheme — including delays in
preparing POIRs which led to POL seeking extensions of time from the Working
Group. There were a number of reasons for this.
159. First, as I have already mentioned, there were more applications to the scheme
than had originally been anticipated and an increasing number of CQRs started
to make their way to POL for investigation from around the turn of the year.
Second, and perhaps unsurprisingly, it took some time for POL’s in-house
investigative team to get into a rhythm of identifying what steps were needed to
investigate a complaint, carrying out those steps, and preparing the resulting
POIRs. This was particularly apparent in cases where the issues were complex
or there was a lack of evidence due to a case being very old. Third, and in a
similar vein, it took longer at first for the POIRs to be reviewed and cleared for
release to SPMs and Second Sight. This was a new process for POL, and I recall
that at this stage in 2014 Rodric Williams was endeavouring to read each draft
POIR prior to release alongside his other work.
‘17 Consent Order: WBON0001667
Page 95 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
160. The speed at which complaints progressed through the Mediation Scheme was
a significant issue generally (not just for POL, and not just at the 13 March 2014
meeting), and I recall that delays in completing CQRs, POIRs and CRRs was a
regular topic of discussion at Working Group meetings. For example, this is
reflected at §6.3 of POL00026672 (minutes of the Working Group meeting on 10
July 2014) about which I am asked at Q42.1 of the Request. Whilst I do not recall
that particular discussion, §§6.3-6.4 accord with my recollection, which is that:
the depth and standard of POL’s investigations into SPMs complaints was
considered by the Working Group to be good; there was nothing in the
suggestion that some SPMs apparently made that POL was deliberately holding
up the progression of cases through the system; and there were various reasons
for delays, not all of which lay at POL’s door. Nevertheless, the issue of delays
and backlogs on all sides persisted throughout the scheme and I have no doubt
that it caused frustration for all concerned, and particularly SPMs.
161. I am asked (by Q40) to set out my recollection of Second Sight's criticism of the
quality of POL’s POIRs at the Working Group meeting on 17 April 2014
(POL00026652). Whilst I don’t recall the specifics of the discussion at that
meeting, I do recall that (i) Second Sight’s complaint came as something of a
surprise to POL, as they had previously been positive about the form and content
of POL’s early POIRs, and (ii) it was not obvious to POL at first why Second Sight
thought the POIRs were deficient as Second Sight did not clearly articulate their
concerns.""®
118 WBON0000821, attaching WBON0000822.
Page 96 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
162. For my part, I felt there was some merit in the suspicion expressed by Rodric
Williams that Second Sight’s complaint about POL’s POIRs stemmed, at least in
part, from the fact that they were struggling to keep up with their own caseload
at that time."'® However, I could also see that there was room for improvement
in some of the early POIRs. A number of them did not reach clear conclusions
(the original format for the POIR did not have an executive summary which Sir
Anthony Hooper later asked to be added to each report) and did not tie POL’s
conclusions closely enough to the underlying evidence. To help POL overcome
this, POL asked the WBD team to step up its involvement in drafting the POIRs
(especially the executive summaries), so as to make them clearer, more strictly
focused on the issues raised by the SPMs, and with tighter referencing to the
evidence underpinning them. I recall that this was in line with feedback from Sir
Anthony Hooper.'2°
(iii) Emerging concerns about Second Sight (Q35, Q41, Q42.2 to Q44)
163. The documents to which the Inquiry refers at Q41, Q42.2 and Q43 broadly relate
to concerns which I expressed about Second Sight in early 2014. To
contextualise those matters, I explain here that I developed a number of concerns
about Second Sight’s role and the work they were carrying out as the Mediation
Scheme progressed. These concerns were to some extent interconnected, and
they were (so far as I was aware) shared by POL and other of professional
advisers to POL including Linklaters, who had been engaged to advise POL on
the terms of the Subpostmaster Contract (the “SPMC”) and the prospect of POL
119 WBONO000821.
120 See for example my email to the team of 21 May 2014, providing feedback on the approach
that was now being taken to drafting POIRs: WBON0000404.
Page 97 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
having any contractual liability to SPMs by reason of the matters advanced by
them in their complaints. I summarise my concerns as follows:
164. First, and most significantly, I was concerned that Second Sight were
insufficiently rigorous in their approach. They frequently appeared unwilling or
unable to drill down into the detail of SPMs’ complaints or POL’s processes in
order to get to the root cause of the accounting shortfalls about which SPMs were
complaining. This was reflected in a lack of cited evidence and analysis in their
reports, and I was concerned that this approach would create unrealistic
expectations on the part of SPMs and ultimately make it difficult to reach
settlements or even achieve some sort of closure in scheme cases, which, after
all, was the whole point of the Mediation Scheme.
165. This problem also manifested in Second Sight being frequently unwilling to
engage with POL in order to obtain further factual detail relevant to their
investigations. For example, it proved difficult to get Second Sight to engage
constructively with POL about the content of the Horizon Factfile with a view to
refining and adding to it to assist them in preparing their Part 1 Briefing Report.'2"
To provide another example, I recall an occasion on which over a dozen subject
matter experts from different teams at POL were brought together for an in-
person meeting at which Second Sight were invited to ask any questions. I
believe that this was while Second Sight were working on their Part 2 Briefing
Report and that the meeting was designed to enable them to ask any questions
121 See for example WBON0000820; see also WBON0000847 where I outline my emerging
concerns in relation to Second Sight’s work (including their lack of interaction with POL at point
7).
Page 98 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
they wanted of POL.'?2 My recollection is that Second Sight refused to engage
and closed the meeting without asking any questions.
166. Second, I was concerned that Second Sight were not able to keep up with their
workload or produce properly considered CRRs in a timely way. As I have
explained above this issue was by no means confined to them, but it appeared
to me that they found it particularly difficult to get on top of their workload even
after POL funded a third forensic accountant, Chris Holyoak, to join their
investigation team. By the time POL decided to close the Working Group in 2015
(which was shortly after it had produced the last of its POIRs), Second Sight still
had around half of their CRRs to produce with the result that the Working Group
had not yet voted on the suitability for mediation of these cases. '7°
167. Third, I became concerned that Second Sight were acting beyond the proper
scope of their instructions, seeking to investigate and opine on matters in which
they had no expertise. It is important to bear in mind that their expertise was as
a firm of forensic accountants, and I (and those instructing me at POL) felt that
they should have stayed within the bounds of their expertise and remained
focused on seeking to identify what had been the root cause of a particular
accounting shortfall in a given case. By way of example, in their draft Part 2
Briefing Report they commented extensively on the ‘fairness’ of the SPMC. I,
those instructing me at POL, and Linklaters, all considered that this was beyond
the scope of their expertise. '24
122 POL00220159.
123 The approximate number of CRRs which had not been provided by the beginning of March
2015 can be seen from this document (updated in May 2015), column Al of which shows the
date on which each CRR was received by WBD (from POL), with blanks where the CRR had
not yet been provided: WBON0000413.
14 POL00021814; POL00207175.
Page 99 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
168. Fourth, I became increasingly concerned that Second Sight’s investigations and
reports were one-sided. This manifested in a number of ways. For example, and
with reference to Q42.2, although I have no specific memory of the meeting which
POL00026672 records, I am able to say that the phrase “hardening of CQRs”
referred to a process by which Second Sight were assisting SPMs to refine CQRs
so as to better articulate their complaint. I recall that it emerged that Second Sight
had been speaking to some of the SPMs before and/or shortly after they
submitted their CQR, in order to help the SPM clarify (or “harden”) their CQRs
and to suggest further information which the SPM should put into it. In some
respects, this was helpful, as the clearer the issues the easier it was for POL to
properly investigate the complaint and produce a clearer and more definitive
POIR. However, I was also concerned that in adopting this approach, Second
Sight were helping the SPMs formulate their complaints and so were losing their
impartiality, and that they were starting to investigate complaints before POL had
had an opportunity to comment.
169. To provide another example, on reviewing many of Second Sight’s CRRs, and
their generic or ‘thematic’ reports, I formed the view that they were frequently
ignoring or marginalising the evidence that POL had provided in response to the
concerns raised. With reference to Q41, this is why I expressed the concern in
my email dated 6 March 2014 that Second Sight appeared to be “inherently
biased” against POL (POL00074462). The respects in which I felt that Second
Sight had ignored information provided by POL or had otherwise failed to
evidence their report in case M001 are summarised in that email.'?5 I recognise
now that the expression “inherently biased” may have been a bit too strong, but
125 As another example, see WBONO0000853.
Page 100 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
I did feel that there was a prevalent lack of balance in the CRRs. I should add
that at the time I expressed that view in March 2014, my views were only
preliminary, though I continued to be concerned about the quality of Second
Sight’s work as time went on.
Early draft thematic report
170. By Q43 of the Request, I am asked to explain various comments which I made
on an early draft ‘thematic’ report which Second Sight provided to the Working
Group in March 2014. I reviewed it as I expected that it would give an early insight
into Second Sight’s approach and thinking, and I wanted to understand whether
they had identified criticisms or areas of concern which had substance.
171. Many of my comments were written with this in mind, i.e. they were aimed at
highlighting gaps in Second Sight’s reasoning or in the evidence underpinning
their (draft) conclusions. They were largely written as questions that could be put
back to Second Sight (rather than questions to POL). So, for example:
171.1.§2.6 of the draft report appeared to be suggesting that SPMs were
impeded in their ability to investigate accounting shortfalls because POL
had control of certain unspecified “back-office accounting functions”. My
comments on that paragraph were intended to draw out the fact that
Second Sight had not explained which “back-office accounting functions”
they had in mind, or why that meant SPMs were therefore unable to
adequately investigate shortfalls (cf. Q43.4). Equally, to the extent that
SPMs were sometimes reliant on POL to provide information to enable
shortfalls to be investigated, Second Sight had not identified from where
they had obtained the information that “requests for investigative support
Page 101 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
or extracts of Horizon data are often refused”. I felt that the source of this
information should be identified by Second Sight in their report and flagged
this in my comments.
171.2.With reference to Q43.7, my comment above Section 5 reflects that I
thought this section was incomplete and unbalanced. I was not saying that
an SPM would only request transaction data where they had made a
mistake. Rather, my comment was intended to convey that since in some
cases it would have been an error by the SPM that caused the loss, the
SPM's knowledge of what had happened on the ground (who they had
served, when, what steps they had taken, etc) was likely going to be an
important part of the picture. I felt that the draft report did not reflect on the
relevance of this to the points that were being made in Section 5.
171.3.With reference to Q43.8, my comment above Section 6 should be read in
the context that, at that stage, I was not aware of any functionality by which
transactions could be entered into branch accounts other than by the SPM
themselves (as I explain further below at §§202 ff). I therefore thought that
this section referred the automatic transaction reversal process which was
the subject of Spot Review 1 and which I have discussed above at §§82-
84); see §6.2. My view at the time was that POL had adequately explained
that process — and in particular, why it need not result in any loss to an
SPM if it was correctly followed — in the context of Spot Review 1. Hence
my comment above Section 6, that Second Sight needed to do more to
explain why such ‘reversed’ transactions could be regarded as a root
cause of SPM losses.
Page 102 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
171.4.The same point applies to Q43.10 and my comment on §7.3. That
comment reflects that in Spot Review 1, it had been acknowledged by POL
that the readily available Horizon data logs did not clearly differentiate
between system-generated reversals and manually inputted transactions.
However, Spot Review 1 went on to conclude that the system made it
possible for SPMs to avoid any difficulties that might arise out of this,
because the fact that a transaction had been automatically reversed
should be revealed by the disconnect and recovery receipts that would be
printed physically in branch. It seemed to me that §7.3 of the draft report
did not take account of this aspect and my comment reflects this.
171.5.Similarly, in relation to Q43.11, my comment (“Need to explain the
relevance of the quote [sic] passage below to this issue”) reflected that I
thought Second Sight needed to explain why they quoted the passage of
the Helen Rose Report that appeared at §7.5. The quote (which made the
above point, that certain Horizon data logs did not clearly distinguish
automated reversals from SPM-input transactions) appeared immediately
below a sentence that read “/the] misuse of User IDs for system generated
transaction reversals appears to be inconsistent with various assurances
and evidence provided by Post Office”. That concerned me, because it
appeared to link the subject matter of the Helen Rose Report to the issue
of transactions being manually entered into branch accounts using SPMs’
IDs and so suggested that the automatic transaction reversal process was
somehow evidence of remote access. As far as I was concerned and knew
at the time, any such suggestion was misconceived because (i) the
automatic transaction reversal process had nothing to do with remote
Page 103 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
access, and (ii) POL had acknowledged the point made about the
presentation of data relating to system-generated transaction reversals.
172. Whilst I was critical of Second Sight’s analysis of the points referred to above,
there were points in the draft report which I thought POL needed to consider
further (cf. Q43.2). For example:
172.1.My comments on Section 4 of the report indicate that I felt that POL needed
to confirm / clarify whether some of the factual statements made by
Second Sight were correct. For example, the description at §4.2 of the
process which SPMs were apparently required to follow in order to ‘Rem
in’ National Lottery to Horizon and the statement at §4.6 that POL failed to
advise SPMs (through either the Helpline or training) that they needed to
reconcile their stock figures for National Lottery products on their Camelot
and Horizon terminals on Thursday mornings instead of at 17:30 on
Wednesday evenings.
172.2.There were factual assertions in §§5.3-5.4 which were new to me and
which I flagged for POL to look into (namely, the assertion that Horizon
only produced a daily record of the aggregate number of value of debit and
credit card transactions without providing a breakdown of those
transactions).
172.3.Where there were points which Second Sight raised which I thought might
have a bearing on criminal prosecutions, I highlighted in my comments
that POL should obtain Cartwright King’s views. For example, and with
reference to Q43.6, at §3.16 of the draft report Second Sight appeared to
be alluding to the idea that the Helpline had advised SPMs to intentionally
Page 104 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
submit false accounts (or had given advice which caused them to do so).
My (lay) view was that it could be relevant to a prosecution if an SPM were
to say that they were told to submit false accounts, so I flagged that POL
needed Cartwright King’s input. I do not know whether POL obtained
advice from Cartwright King on this point.
173. With reference to Q43.3 and Q43.5, these were matters which went to the
contractual relationship between POL and SPMs. I did not consider that these
were matters that merited a wider-ranging investigation by POL at that stage,
because (i) they were well outside the scope of Second Sight’s expertise to
comment on, and (ii) they were outside the scope of the issues then being
considered, namely whether the Horizon system and associated processes were
responsible for the shortfalls about which individual SPMs within the scheme
were complaining.
174. Finally, in relation to Q43.9 of the Request, my comment on §7.2 reflects the fact
that I considered the language of “ghost” transactions to be rather sensationalist
and in that sense, “dangerous”. Taken in isolation, the sentence “/n some
instances these ‘ghost’ transactions appear to have contributed to shortfalls for
which the relevant Subpostmaster was later held accountable” suggested that
SPMs had wrongly been held liable for improperly inputted transactions, which
were not initiated or approved by them, and which they had no way of
discovering. In fact, Second Sight went on in the following paragraphs to rely only
on the automatic transaction reversal process in support of this sentence. As I
have already explained, this process did not enter transactions ‘improperly’ (but
rather did so in accordance with the intended design of the system) and it did not
enter transactions without the SPM having the means to discover this. Given that
Page 105 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
this was a draft of a report which could ultimately end up in the public domain, I
was concerned by the language used and thought that POL should obtain
Cartwright King’s advice on the implications of this paragraph remaining,
unamended and unedited, in any final report. Similar points applied in relation to
the statement in §7.4, that POL had “misuse/d] User IDs for system generated
transaction reversals”.
Discussion on Second Sight’s draft Part 1 Briefing Report
175. In answer to Q44 of the Request, I regret that I am unable to recall the discussion
referred to in POL00026662, concerning Second Sight’s draft Part 1 Briefing
Report.
(iv) Redactions to / provision of the Helen Rose Report (Q32, Q49)
176. I deal here with Q32 and Q49 of the Request, which raise related questions
arising out of attempts made by Mediation Scheme applicants to obtain
disclosure of an unredacted copy of the Helen Rose Report in the context of the
scheme. This began with the email dated 7 April 2014 at the start of the chain in
POL00116487 (cf. Q32), by which an applicant's solicitor sought disclosure of a
“full, final and unredacted” copy of the report of behalf of her client, the applicant
in question having already received a redacted version by way of post-conviction
disclosure from POL (on the advice of Cartwright King). As time went on, an
increasing number of applicants (including applicants who had not received the
redacted version by way of post-conviction disclosure) made similar requests for
unredacted copies of the Helen Rose Report in their CQRs; this was the
background to my email of 17 June 2014 contained in POL00129392 (cf. Q49).
Page 106 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
177. In order to understand these requests and POL’s response to them it is
necessary to appreciate that the significance of the Helen Rose Report was
dependent on the context. First, and as I have identified above at §§86-89, the
report was undoubtedly important in the context of historic prosecutions where
Cartwright King had advised that Gareth Jenkins had given misleading testimony.
I was aware that for this reason, post-conviction disclosure of the report had been
given in a number of cases (such as that of the applicant referred to in
POL00116487). However, whether POL was required to disclose the report to a
convicted SPM was not a Mediation Scheme matter, but rather something to be
managed outside of that process by Cartwright King as POL’s criminal solicitors.
178. Second, and quite apart from the issues relating to Gareth Jenkins’ evidence in
criminal cases against SPMs, the Helen Rose Report contained criticisms of the
automatic transaction reversal process in Horizon Online. Specifically, as
explained above at §84, the report criticised the way in which Horizon presented
data relating to automatically reversed transactions, in that it failed to make clear
that they were system-generated as opposed to having been manually
undertaken by the SPM. These criticisms may have been relevant in a case
where the applicant complained that the automated transaction reversal process
had caused them to suffer a particular loss. However, my recollection is that there
were few Mediation Scheme cases where the applicant did make such a
complaint. Applicants who sought disclosure of the Helen Rose Report in their
CQRs therefore tended to do so in generalised terms, speculating that the
unredacted report provided evidence (i) of some form of remote access capability
that POL was said to have, or (ii) that the system could generate transactions
which were entered into the branch accounts with the SPM’s own User ID
Page 107 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
attached other than in the narrow context of the automatic transaction reversal
protocol, without pointing to any particular transactions that they disputed as a
result. In these circumstances and as I understood matters then, the Helen Rose
Report had little if any bearing on the actual the facts of these applicants’ cases,
and I felt — and POL agreed — that routinely providing copies of the report risked
being an irrelevant distraction from investigating the actual issues raised by each
applicant.
179. The reference to “downplaying” / “minimalizing” the importance of the report in
POL00129392 (cf. Q49) should be understood in that light. It was not a comment
about the importance of that document to the issues surrounding Gareth Jenkins’
evidence and which were being separately addressed by Cartwright King. From
my (civil lawyer) perspective, it was intended to convey that in the context of the
Mediation Scheme the report was believed to be something of a red herring and
that POL would not usually need to enter into debate about the report or its
contents in Mediation Scheme cases (an email dated 31 July 2014 from Andy
Pheasant, a WBD solicitor working under my supervision, provides an example
of what I regarded as the right approach’”°). Cartwright King, separately and for
their own reasons connected with the criminal process, were concerned about
POL routinely providing the Helen Rose Report to Mediation Scheme applicants
who were not entitled to a copy by way of post-conviction disclosure. As such,
my email in POL00129392 is drafted in terms which reflect their advice, and it
should be read against that context.
126 WBONO000888; see also the attachment: WBON0000889.
Page 108 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
180. I add that the sentence which recommends that POIRs should, so far as possible,
‘minimalise’ or ‘ignore’ the Helen Rose Report should be taken together with the
next sentence: “/f the investigation team need guidance on how to address any
HR Report related questions, I suggest that they (or the lawyer here at BD)
addresses these directly with CK”. I recall that it was part of the criminal lawyers’
remit to keep under review the question of whether the report needed to be
disclosed to individual applicants who had convictions (albeit that if this
happened, it would take place through prosecution rather than Mediation
Scheme channels). The latter sentence reflects this.
181. As regards Q32 and POL’s approach to redacting the Helen Rose Report, on
receiving Priti Maru-Singh’s email dated 7 April 2014 I sought Cartwright King’s
input. This was because it was a query relating to disclosure that had been given
post-conviction in a criminal case: POL00116487. I do not specifically recall the
conversation I had with Simon Clarke about Priti Maru-Singh’s request for an
unredacted copy of the report, but it is evident that he advised that the rationale
for the redactions which Cartwright King had applied was that they were
necessary to comply with data protection legislation (and in one case, to remove
an assertion of LPP which Cartwright King believed was wrong and could have
been confusing). His position was that the redactions should be maintained.
182. Leaving aside the incorrect reference to privilege in the header, I could see that
all the other redactions related to (i) the SPSO’s location and branch code, (ii)
the SPM’s user ID, and (iii) the names of individuals, namely Gareth Jenkins (the
Fujitsu employee) and Helen Rose (the Report's author). I agreed with Simon
Clarke that this data constituted personal data such that the prima facie position
was that it should be redacted. For the avoidance of doubt, I had no view on (and
Page 109 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
it was not my role to advise on) how POL's criminal law disclosure duties
interacted with the requirements of data protection law; this was for Cartwright
King to advise on and I took at face value their assessment that the redactions
did not put POL in breach of its prosecutorial duties. I therefore recorded the sum
total of my and Cartwright King’s views (together with those of Rodric Williams,
to whom I also appear to have spoken though I have no specific memory of this)
in my email to Belinda Crowe dated 8 April 2014, to which I am referred by Q32.
183. For completeness, on 9 April 2014 I spotted that the redacted version of the
report sent to Priti Maru-Singh’s client by POL / Cartwright King did not appear
to include the appendix found in the original version sent to me in 2013. I emailed
Simon Clarke to ask whether Cartwright King had disclosed the appendix (which
contained the relevant credence data, Fujitsu transaction logs and other
information relevant to the transaction reversal process) as part of post-
conviction disclosure and if not, why it had been omitted.'2” He responded that
day that he had not seen the appendix prior to my bringing it to his attention, but
in any event in his view the appendix was not disclosable in the ‘criminal arena’.
He explained that the reason for disclosing the Helen Rose Report was its
potential to impugn Gareth Jenkins’ credibility as a witness, and the appendix did
not speak to that issue.'76
184. I proceeded to review the appendix and concluded that it should be provided to
Priti Maru-Singh and her client together with Spot Review 1, notwithstanding
Simon Clarke’s view that it was not relevant from the criminal law perspective. In
particular, I thought that it might be beneficial in helping Priti Maru-Singh and her
127 WBONO000828.
128 WBONO000834.
Page 110 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
client (i) to understand the subject matter of the Helen Rose Report, and (ii) to
reassure them that there was not extensive additional or supplemental material
that POL was withholding (as Priti Maru-Singh appeared to believe). I sent
Belinda Crowe an email to this effect on 14 April 2014, attaching the appendix
and Spot Review 1 (which I redacted to remove personal data in line with the
approach taken to the main body of the report).129
(v) Provision of Officers’ Reports (Q46)
185. At around the same time (i.e. in early 2014 when the process of investigating
SPMs' complaints and producing POIRs was getting going) an issue arose as to
the extent to which POIRs could refer to, and attach, reports that had previously
been compiled by POL investigation officers in the course of investigating
suspected criminality in a branch (“Officers’ Reports”, also sometimes called
“Offender Reports”). Such reports often contained information about what had
happened in a branch, which was useful in understanding and responding to the
complaints of convicted SPMs who were within the Mediation Scheme. This was
particularly true in some of the older cases where few other contemporaneous
records remained.
186. From a purely civil perspective, I thought that if an Officer’s Report contained
useful contemporaneous material which informed POL’s response to an
applicant’s complaint in its POIR, then the report could be provided to the
applicant and Second Sight as supporting evidence. Although there was no duty
of disclosure in the context of the Mediation Scheme, my general view was that
128 WBONO0000838.
Page 111 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
providing documentary evidence in support of POIRs where possible would
make for a more credible response than simply responding by way of assertion
in the POIR itself. In turn, I thought this would help applicants to better
understand POL’s position and would be more likely to bring about closure in
Mediation Scheme cases.
187. However, these were not the only considerations, because POL’s criminal
lawyers raised concerns that, from their (criminal law) perspective, provision of
the Officers’ Reports could cause difficulties. From reviewing my emails, the
criminal lawyers’ concerns first came to my attention on or around 7 April 2014
when Jarnail Singh (POL’s internal criminal lawyer) forwarded an email from
Andrew Bolc of Cartwright King to me and Rodric Williams.'®° Andrew Bolc’s
email referred to case M006, in which (it seemed) the draft POIR had proposed
to attach the Officer’s Report from the original criminal investigation into the
applicant. Andrew Bolc stated that Harry (Bowyer, in-house counsel at Cartwright
King) had “worries about these documents being disclosed let alone without
being redacted” because they were “prosecution working document[s]’ and
because of data protection concerns. Cartwright King sought a decision from
POL as to whether Officers’ Reports (i) should not be disclosed in any
circumstances, (ii) should be disclosed subject to redactions, or (iii) should be
disclosed in full “accepting any consequences that follow’."*'
188. I did not think that Officers’ Reports (being a record of an internal investigation)
would ordinarily meet the test for legal professional privilege in civil proceedings,
but I did not know what the position was in the criminal sphere; in particular, I did
13° WBONO000825.
131 WBONO000825.
Page 112 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
not know what was meant by “prosecution working document[s]’. I therefore
responded to Andrew Bolc asking “/flrom a criminal law perspective, what [the
consequences are] of disclosing a document that had previously been withheld
in a prosecution on the grounds that it is a prosecution working document” .192
189. Harry Bowyer (rather than Andrew Bolc) responded on 8 April 2014 and it is worth
setting out his advice in full:
“If we are to be serving these documents then it should be an informed
decision of our mutual client to do so as there may well be
consequences.
Please forgive me if I appear to be teaching my grandmother to suck
eggs in the following paragraphs but I will be grateful for the same when
you teach me civil disclosure!
The documents that we are concerned with are the officers’ reports.
These are prepared at a very early stage of a prosecution and are
intended to set out the facts and background of a case in order that a
decision to prosecute might be made. This is necessarily at a stage when
the investigation is far from complete and will often contain conjecture
and opinion that will subsequently be proved wrong or inflammatory. I
was reviewing a case yesterday where the officer was wondering
whether the suspect was taking the fall for her daughter when the
daughter was, in his view, more than likely to be involved.
They will contain criticism by the officer of POL procedures and
suggestions for putting them right — whether these are acted upon history
seldom relates.
They also contain, in many cases, operational material that shows how
these cases are detected and the investigational resources that are
available to POL. This is not something that should be released into the
public domain lightly — especially where the audit is intelligence led.
There are certain of these documents where information is revealed, no
[sic] relevant to the case, which may be commercially sensitive or
embarrassing to our client. The case of Walters M006 has an example
where the officer raises the concern that there were no checks made on
spoiled postage slips to see if they were bogus or not. We do not know
whether this has been fixed or even applies today.
132 WBONO0000825.
Page 113 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
The final area of concern is that a substantial minority of these
applications contain complaints about the behaviour of our investigators.
These documents give the telephone numbers and other contact details
of the officers who compile the reports which presumably may well find
their way into the hands of those who have a long held animus against
them. In a world governed by the Data Protection Act we should think
extremely carefully before sending documents out unredacted even to
this extent.
These documents are seldom, if ever disclosed to the defence as they
are not the primary evidence and are a prosecution working tool. If they
contain information that the defence should have we usually serve it in
some other way — either by statement, documentary exhibit or a
disclosure note which will say that, “Post Office Limited are aware
that...... ,
This information is and documentation is, in the main, POL’s. Where it is
POL’s documentation and POL’s information there is nothing to prevent
its disclosure by POL (subject to the above) even where we have made
the decision not to disclose the document in the criminal proceedings.
This is why we have asked for clarification as to what POL wishes to do
and the options are: 1) Disclose unredacted, 2) Disclose redacted copies
or 3) Do not disclose.
We need a consistent approach or people will notice that we are serving
them in some cases and not in others.”"°
190. Jarnail Singh responded on the chain expressing the clear view that: “Having
read Counsel Bowyers advice on disclosure of the investigation officers report In
my view the business need to take the view not disclose to such documents at
alf’ 154
191. It was not immediately clear to me why Jarnail Singh thought there should be an
absolute prohibition on disclosure of these documents. My understanding of
Harry Bowyer’s email was that whilst Officers’ Reports were generally not
disclosed in criminal proceedings, he did not think there was any inherent reason
183 WBONO000825.
134 WBONO000825.
Page 114 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
why they could not now be disclosed, at least in part, if it was thought appropriate
to do so. They did not, for example, contain legal advice, and he had appeared
to confirm that their disclosure would not automatically have consequences in
the criminal proceedings in which they had originally been withheld. At the same
time, however, Harry Bowyer identified a range of material which might appear
in such reports which should not be disclosed, e.g. sensitive information about
POL’s operational and investigative techniques. Since the main reason for
disclosing such reports was in order to evidence specific points made in POL’s
POIRs, I tentatively thought that the way forwards was to allow disclosure, on a
case by case basis, of those parts of the reports which were germane to the point
that POL was making (and which did not reveal the type of sensitive information
about e.g. investigative techniques which Harry Bowyer had in mind).
192. I relayed the understanding I had derived from Harry Bowyer’s email, and my
advice, in POL00061369 (my email to Belinda Crowe and Rodric Williams of 17
April 2014). For the avoidance of doubt, and with reference to Q46.2, my
statement that “/ cannot see that this document would attract legal privilege as it
is an investigation document and not a document prepared for the purposes of
litigation” is reference to my view that Officers’ Reports would not be privileged
from disclosure in civil proceedings (i.e. so this was not a material consideration
in deciding whether such documents should generally be withheld). I also relayed
my understanding of the advice from Cartwright King that "This document is
typically not disclosed through the prosecution process as it is part of the
prosecution working papers and therefore, I understand, it is usually exempt from
disclosure". Jarnail Singh (POL's in house criminal lawyer) was copied to the
email and I asked him to correct me if I had got this wrong. My reference later
Page 115 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
on in that email to “prosecution privilege” is a reference to what I had understood
from Harry Bowyer, namely, that Officers’ Reports were usually exempt from
disclosure in criminal proceedings. As I have explained, I thought that the net
effect of all this was that there should not be a blanket ban on providing relevant
parts of POL’s Officers’ Reports to SPMs, but that rather they should be provided
if and to the extent that they contained relevant information that was not
privileged or sensitive (such as information which revealed details of POL
investigative techniques).'°°
193. Angela Van Den Bogerd agreed with the approach I suggested and gave
instructions to that effect on 22 April 2014:
“We do refer to the officer's report in case MO54 and in this instance
using this report does in my view make for a more conclusive case.
Therefore my view is that this needs to be addressed on a case by case
basis as you suggest but with a presumption against disclosure unless
absolutely necessary.
Kath, Shirley — please ensure that: if you wish to use an investigation
officer's report as a supporting document that you flag this to BD When
you send them the report so that they can advise accordingly.”°>
194. Jarnail Singh continued to advocate for a blanket ban on the basis that Officers’
Reports were “a prosecution working toof and that relevant information
contained in such a report would normally be disclosed (in criminal proceedings)
by some other means such as a witness statement.'>” That proposal was of no
help in the present context (i.e. of a mediation process) so I asked Jarnail Singh
135 For completeness, as an email exchange with Brian Altman QC in July 2016 shows, it was only
some years later that I learned that documents which revealed information about investigative
practices are not in fact privileged from disclosure in the criminal law context, unless and to the
extent that that information attracts Public Interest Immunity: WBON0000443. See further
below, §408.
136 POL00061369.
137 POL00061369.
Page 116 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
to clarify whether there was any particular reason why, as a matter of criminal
law, Officers’ Reports could not be disclosed per se."** I found his response dated
23 April 2014 very difficult to follow (and I am not at all sure what he meant by “/
am less concerned with the fact that investigation report is not signed statement
but more concerned with the potential content of the report which may be
potentially damaging to the POLS interest’).'*° Ultimately, it did not change the
views I had expressed in my earlier email, and I did not understand it to change
Angela Van Den Bogerd's instructions since she did not respond further to this
chain altering the directions she had given.
195. I can see from reviewing my email records that the approach I had outlined, and
which was seemingly approved by Angela Van Den Bogerd, was adopted at first.
For example, on 27 April 2014, Angela Van Den Bogerd reviewed the draft POIR
for case M019 and asked: “surely we can use some of the information we
gathered in the Security led investigation to crystallise our conclusion in respect
of what happened in this case?” In the event, in the particular circumstances of
the case I did not consider that the Officer's Report added anything to the
material already relied on, so I recommended that the report did not need to be
referenced.'4°
196. Matters subsequently came to a head, and POL’s instructions on how to
approach the issue of disclosure of Officers’ Reports changed, when a case
arose where I felt that the report did need to be referred to (along with another
document from the prosecution file). This was case M029. I raised this case with
138 POL00061369.
139 POL00061369.
140 WBONO000848.
Page 117 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Rodric Williams on 6 May 2014, explaining that I felt we needed to disclose the
documents. Rodric Williams responded the same day saying that the decision
needed to be run past Cartwright King.‘4"
197. I therefore emailed Martin Smith of Cartwright King to seek his views (Jarnail
Singh in copy), identifying that the reason for the proposed disclosure was that
the documents were “important to prove the conclusions reached in the POL
report as there are no alternative documents on which to rely” (POL00046216).
Martin Smith responded on 7 May 2014, acknowledging that whilst Cartwright
King had “advised that as a matter of principle investigation and offender type
reports should not be disclosed ... there will be cases in which it is felt that there
is no alternative other than to disclose these”. He advised that in such cases
Officers’ Reports should be “appropriately redacted” and it was agreed that he
would identify the redactions for case M029 (POL00046219). This was in line
with what I understood to be the approach following the exchange in
POL00061369, i.e. of considering disclosure subject to redactions on a case-by-
case basis.
198. At this point (on 8 May 2014) Jarnail Singh intervened, escalating the matter to
Chris Aujard (then POL’s General Counsel) in the following terms:
“As I understand it, POL has been advised by senior counsel that
investigation and offender report should not be disclosed. It is of course
matter for POL to make a decision whether [sic]
to accept this advice or not and of course it would be open for POL to
decide to discourse such documents. I personally would be unhappy for
such documents to be disclosure for reasons set out in counsel Harry
Bowyers advice note.
141 WBONO000849.
Page 118 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Given the email correspondence between Bond Dickinson and
cartwright King, I would be grateful if I could be informed whether POL
has made a decision or Bond Dickinson are proceeding along the
disclosure route without POL having made a decision.”'42
199. Also on 8 May 2014, Martin Smith emailed me proposed redactions in case M029
(POL00046219). I cannot recall whether I considered the specifics of those
redactions, as POL was by then reconsidering the approach to disclosure of
Officers’ Reports. Later that day, Rodric Williams emailed the following (revised)
instructions on how such reports should be dealt with in POIRs:
“Having discussed this with Chris and Jessica, the protocol for the use
of ‘Officer Reports’ (or as otherwise described) by Project Sparrow
investigators when responding to individual complaints is:
1. The report is NOT to be exhibited OR expressly referenced in Post
Office's formal response to a complaint.
2. It can be used by the investigator to help them understand what
happened in a particular case, and to identify other documents relevant
to the case (e.g. transcripts of interviews, branch account records etc).
3. If the report is the ONLY source document still available, the
investigator can repeat material from the report (provided it is not legally
privileged), but CANNOT cite the report as a reference.
4. Any challenge received about the source of a Post Office statement
made from the report must be referred to Chris.”"*9
200. Accordingly, the issue of Martin Smith’s proposed redactions to the prosecution
documents in case M029 fell away.
201. I forwarded POL’s revised instructions to Angela Van Den Bogerd and the POL
investigating team on 9 May 2014, acknowledging that the approach “may cause
a few headaches as the officer's reports are key doc”, but that it “represent[ed]
142 WBONO000850.
143 WBONO000403.
Page 119 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
the firm view for Chris on the way to proceea”.'** On 21 May 2014 I sent the
following update to the WBD team working on the Mediation Scheme, which
reflected my understanding of the instructions to which we were now required to
work:
“Officer's reports:
o Post Office has taken a decision that it will not be disclosing
‘officer's reports’ or other similar reports from the Post Office
security team — sample attached.
© This is because of various criminal law / prosecution issues.
o We can use information in the officer's report in Post Office's
investigation report (indeed you may copy the information word-
for-word) but cannot refer to the officer's report or disclose them
to Second Sight.
o In some places, that will mean making un-evidenced statements
which are not supported by any document however POL is happy
with that risk.
o If, further down the line, we receive a complaint from SS/the
Applicant that the officer's report has not be disclosed, please
escalate to me.
o Note: sometimes these are also called offender's reports or
investigation reports. If you are unsure, speak to me.”45
(vi) Provision of information concerning the Balancing Transactions
functionality (Q45)
202. Q45.1 and Q45.2 of the Request ask me to set out my beliefs as to (i) the extent
to which Fujitsu had inserted additional data into branch accounts without the
144 WBONO000851.
145 WBONO000404.
Page 120 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
knowledge of SPMs in Old Horizon, and (ii) my understanding of the security
measures Fujitsu had in place regarding the use of remote access, at the time of
the email chain in FU00087119."4° This chain refers to enquiries arising out of
the emergence of an email dated 23 October 2008 between Andrew Winn (a
Relationship Manager in POL’s Financial Service Centre) and Alan Lusher (a
Contracts Adviser in POL’s Network Support Team) (“the Winn/Lusher email”;
POL00117650).
203. The Winn/Lusher email came to my attention because it was referred to in an
email by Steve Darlington (Howe+Co’s Finance Director), which was forwarded
to Belinda Crowe by Second Sight (and by Belinda Crowe to me) on 8 April
2014.'47 I was not aware of the email before this point.'4® Angela Van Den Bogerd
managed to obtain a copy of the email and on 14 April 2014 she circulated the
chain, though I do not know from where she obtained it.14°
204. The Winn/Lusher email stated:
“The only way POL can impact branch accounts remotely is via the
transaction correction process. These have to be accepted by the branch
in the same way that in/out remittances are i guess. If we were able to
do this, the integrity of the system would be flawed. Fujitsu_have the
ability _to_impact branch records via_ the message store but have
extremely rigorous procedures in place to prevent adjustments being
made without prior authorisation - within POL and Fujitsu” (emphasis
added).15°
146 For the avoidance of doubt, I was not copied into the email from Sean Hodgkinson (Deloitte) or
Pete Newsome (Fujitsu) of 19 May 2014 and do not believe I have seen them prior to receiving
the Request. The absence of these emails from my firm’s file supports this view.
147 WBONO000826.
148 Nor, it seems, were others in the chain including Belinda Crowe and Angela Van Den Bogerd:
WBONO0000827.
148 WBONO0000835.
150 POL00117650.
Page 121 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
205. I was not aware of the functionality referred to in underline above before I
received the Winn/Lusher email on 14 April 2014, and so I was not cognisant of
the fact that Fujitsu could inject transactions into branch accounting records in
either Old Horizon or Horizon Online (in Horizon Online this was via the
“Balancing Transaction” functionality, however in the early stages of
investigating remote access this phrase was sometimes used to describe
injected transactions in both Old Horizon and Horizon Online because those
instructing me at POL and I were not aware of the technical differences between
the two systems). In order to give the Inquiry a full picture as to how my
understanding of injected transactions developed thereafter, and to contextualise
the advice I gave concerning the provision of this information to Second Sight
and SPMs (cf. Q45.3), it is necessary to set out the enquiries made by POL of
which I was aware as to what changes to branch data the Winn/Lusher email
referred to. I highlight at the outset that these enquiries were complex and
protracted, running in parallel with a significant volume of other work during the
Mediation Scheme. I do not remember the sequence of events well and do not
attempt to be exhaustive, but I have done my best to piece together the key
elements of what happened from my firm’s records.
Initial enquiries
206. Following receipt of the Winn/Lusher email I prepared a set of questions for
Fujitsu.15' Rodric Williams sent questions to James Davidson (Post Office
Account Delivery Executive at Fujitsu) on 17 April 2014 (as shown in
FUJ00087119). James Davidson responded with the following overview,
151 WBON0000837.
Page 122 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
together with more granular responses to the questions raised. This information
was forwarded to me by Rodric Williams on 22 April 2014 (FUJ00087119):
“Summary:
e There is no ability to delete or change records branch creates in
either old Horizon or Horizon online. Transactions in both systems
are created in a secure and auditable way to assure integrity, and
have either a checksum (Old Horizon) or a digital signature
(Horizon Online), are time stamped, have a unique sequential
number and are securely stored via the core audit process in the
audit vault
e Whilst a facility exists to ‘inject’ additional transactions in the event
of a system error, these transactions would have a signature that
is unique, sub-postmaster id’s are not used and the audit log
would house a record of these. As above, this does not delete or
amend original transactions but creates a new and additional
transactions
e This facility is built into the system to enable corrections to be
made if a system error/ bug is identified and the master database
needs updating as a result, this is not a unique feature of Horizon
e Approvals to ‘inject’ new transactions are governed by the change
process, 2 factor authentications and a ‘four eyes’ process. A
unique identifier is created and can be audited for this type of
transaction within HNGX, Horizon would require more extensive
work to investigate as explained below.
1. Can Post Office change branch transaction data without a
subpostmaster being aware of the change? No
2. Can Fujitsu change branch transaction data without a subpostmaster
being aware of the change? Once created, branch transaction data
cannot be changed, only additional data can be inserted. If this is
required, the additional transactions would be visible on the trading
statements but would not require acknowledgement / approval by a
sub-postmaster, the approval is given by Post Office via the change
process. In response to a previous query Fujitsu checked last year
when this was done on Horizon Online and we found only one
occurrence in March 2010 which was early in the pilot for Horizon
Online and was covered by an appropriate change request from
Page 123 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Post Office and an auditable log. For Old Horizon, a detailed
examination of archived data would have to be undertaken to look
into this across the lifetime of use. This would be a significant and
complex exercise to undertake and discussed previously with Post
Office but discounted as too costly and impractical.
3. If not, where is the evidence for this conclusion? See Answer 2
4. If so:
a) How does this happen? See above
b) Why was this functionality built into the system design? To allow for
data to be corrected if there were any defects found in the system
c) Why would Fujitsu need to use this functionality? As above and under
instructions from Post Office Ltd.
d) What controls are in place to prevent the unauthorised use of this
method of access? This is achieved through a number of industry
standard controls (RBAC, 2 factor authentication etc) which are
robustly audited under ISO 27001 / IAS 3402, Link, PCI.
e) When has branch data been accessed in this way in the past? See
above
5. In relation to the Winn/Lusher email:
a) What is "message store"? This is the repository (or database)
where all transactions were written to in the old Horizon system
b) Can this be used to access and change branch records? It can be
used to access the records. Data cannot be changed, but new data
could be inserted into it. Any such inserted data would be tightly
controlled by operational processes explained above.
c) What is the "impact" of this change on branch records? The impact
would depend on exactly what records were inserted.
d) Would the subpostmaster be aware of this change? Yes, via the
trading statement but spm’s are not required to approve the change,
this is provided by Post Office.
e) Why would this method of access be used? To correct errors if a
software defect is identified.
f) What controls are in place to prevent misuse of this method of access?
As above.”
Page 124 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
207. My understanding of the explanation given by James Davidson was that:
207.1.First, Fujitsu had no ability to edit or delete (i.e. change) transaction data
and indeed, any attempt to manipulate or delete such data would be
evident due to the way in which it was held in the audit log.
207.2.Second, Fujitsu could insert a new transaction, but:
(i) This would only occur in exceptional circumstances to correct data.
(ii) It was only with the express approval of POL.
(iii) The insertion would be visible to an SPM as it would show in their
branch accounts under a unique ID which was not that of the SPM or
any of their staff.
(iv) The process had only been used once in the lifetime of Horizon Online.
There was a substantially similar functionality in Old Horizon, but it
would be difficult and expensive (though not impossible) to ascertain
on how many occasions it had been used. ‘52
(v) The existence of a functionality of this type was not unusual in a system
such as Horizon.
208. That this was my understanding is reflected in a draft note (the “April note”)
which I prepared for Rodric Williams and I recommended should be shared with
Second Sight once it had been bottomed out and approved by Fujitsu.1°°
152 I therefore did not know whether this functionality had been used to insert a transaction in Old
Horizon (and if so, how many times), because Fujitsu themselves did not know.
153 WBONO0000845; POL00204068.
Page 125 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
209. Second Sight were seeking POL’s response on the Winn/Lusher email point, so
on 9 May 2014 I recommended that a holding response be sent: (i) confirming
that there was no functionality to edit or delete transaction data; (ii) advising that
“it is possible to input additional transactions into a branch's accounts (e.g. by
way of say a transaction correction), [but that] a SPMR will always have visibility
of these extra transactions as they are shown separately in the branch's
accounts’; and (iii) informing Second Sight that the latter point was being pursued
with Fujitsu and a more detailed note would follow.’ This was, in my view, the
appropriate course; it was prudent for POL to obtain further information about
injected transactions by Fujitsu before attempting to describe them in detail to
Second Sight.15>
210. In the event, the process of finalising the April note with Fujitsu drifted. 15° I do not
now recall the reason why that happened, and I may not have known at the time,
because I believe that I was not sighted on all of the relevant correspondence
with Second Sight and Fujitsu. I later knew (but did not know at the time) that
POL were simultaneously engaging Deloitte to look into similar questions.
211. At around this time, POL needed to consider how to respond to cases within the
Mediation Scheme where the applicant had raised concerns about remote
access in their CQR. By summer 2014, substantial numbers of CQRs were
starting to make their way through to POL and POL needed to investigate them
and produce POIRs in a timely way (as set out above). What information should
154 WBONO0000852.
185 See for example WBON0000854, where I expressed the view to Rodric Williams that POL
needed to understand more about the one occasion the Balancing Transaction tool was said to
have been used in Horizon Online, as well as how difficult it would be to identify whether a
transaction had ever been injected into Old Horizon.
156 See for example WBON0000854; WBONO0000860.
Page 126 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
be provided was ultimately decided on a case by case basis, but as a matter of
principle it was decided that the potential for injected transactions did not need
to be specifically identified and explained where there was no allegation that
particular transactions did not originate in branch'S’ or there was a clear
explanation for the SPM’s allegations (for example, in case MO56 the applicant
had alleged that there was a ‘phantom log-in’ to her account, but on investigation
it was concluded that the issue was caused by her inadvertently leaving a branch
terminal logged on for over a week). '®8
Inquiries following the Project Zebra Desktop Report
212. On 22 August 2014, I received, for the first time, a copy of the Project Zebra
Desktop Report dated 23 May 2014 (POL00028062)."®° For context, I am now
aware that Project Zebra had been commissioned by POL in early 2014,
following advice it had received from Linklaters about POL’s potential liability to
SPMs (which advice is addressed further below at §248). However, my
involvement in that project at the time was very limited, and indeed I was not
even aware it had been commissioned at first. I gained some limited awareness
of it around May 2014, when POL was engaging Deloitte to produce a report and
Rodric Williams asked me to comment on some proposed wording to be included
in its instructions to Deloitte (specifically, five questions which it was proposed
Deloitte should answer, none of which specifically related to injected
157 Often these were in the nature of vague references to the Helen Rose Report which was by
then in circulation within the SPM community (and which did not concern any form of remote
access but rather related to the transaction reversal process), or, as in the case of Mr Rudkin—
which was number M051 in the Mediation Scheme — the allegation was that POL employees
had the ability to pass transaction data into the live Horizon system from the test centre in
Fujitsu’s Bracknell office (which was considered to be unfounded for the reasons given above).
188 POL00307712.
159 WBON0000891.
Page 127 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
transactions).'® After this, I do not believe that I or any of my colleagues at WBD
had any involvement until the Project Zebra Desktop Report was sent to me on
22 August 2014. Even at that stage, my visibility of Project Zebra was limited. I
did not, for example, appreciate that there were any reports beyond the Desktop
Report, and only became aware that there was a further report in February
2016.'®' I also did not know that the project had generated any other reports
beyond these two, until they were put to me in Q88 of the Request.'62
213. The Project Zebra Desktop Report was a 73-page document produced in
response to instructions from the POL litigation team that "POL is responding to
allegations from Sub-postmasters that the "Horizon" IT system used to record
transactions in POL branches is defective and that the processes associated with
it are inadequate (e.g. that it may be the source and/or cause of branch losses)".
The purpose of those instructions was for Deloitte to investigate whether project
documentation, operating policies, and previously undertaken assurance work
appropriately covered key ‘risks’ relating to the integrity of the Horizon processing
environment, including identification of the ‘Horizon Features’ that ensured that
SPMs had full ownership and visibility of movements in their branch accounts.
One of the ‘Risk Areas’ flagged by Deloitte as one of the ‘Key Matters for
Consideration’ was Balancing Transactions (item 4(g) on page 31).:
“g. Branch Database: We observed the following in relation to the
Branch Database being:
A method for posting ‘Balancing Transactions’ was observed from
technical documentation which allows for posting of additional
16° WBONO000856.
161 As is illustrated by my email of 10 February 2016 to Rodric Williams, which stated “/ don’t have
a copy of the Deloitte board report - do you have it?’ WBONOOO0960. I refer to my receipt of
this report further below, at §§299-300.
162 Namely, POL00105635, POL00031384, POL00031391, and POL00029726.
Page 128 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
transactions centrally without the requirement for these transactions to
be accepted by Sub-postmasters (as ‘Transaction Acknowledgements’
and ‘Transaction Corrections’ require). Whilst an audit trail is asserted to
be in place over these functions, evidence of testing of these features is
not available;
[ul
For ‘Balancing Transactions’, ‘Transaction Acknowledgments’, and
‘Transaction Corrections’ we did not identify controls to routinely monitor
all centrally initiated transactions to verify that they are all initiated and
actioned through known and governed processes, or controls to
reconcile and check data sources which underpin current period
transactional reporting for Subpostmasters to the Audit Store record of
such activity;
Security of the Branch Database around the ‘Messaging Journal table’ is
a key area of risk due to the branch transactional data being held on this
table for up to a day before being written to the Audit Store. It was unclear
from the documentation reviewed whether specific assurance work had
been carried out in this area; and
Controls that would detect when a person with authorised privileged
access used such access to send a ‘fake’ basket into the digital signing
process could not be evidenced to exist”. 1°
214. I read the report with an eye on that issue, as I had been provided with it for the
purpose of advising POL on how it should describe Balancing Transactions to
Second Sight (who were by then working on finalising its Part 2 Briefing Report,
the first version of that report having been released on 21 August 2014).1%
215. From my perspective at the time, I did not read the Desktop Report as altering
the understanding of Balancing Transactions I had gained earlier in the year from
the email I had received from James Davidson of Fujitsu (above, §§ 206-208).
Consequently, on 21 October 2014, I recirculated the note I had previously
63 POL00028062.
164 POL00226961.
Page 129 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
prepared, flagging to POL that its content needed to be confirmed with both
Fujitsu and Deloitte before sharing with Second Sight."
216. In the Desktop Report, there are long tables of "Horizon features" in Appendix 2
including three rows (on pages 53 and 54) that reference the possibility that
database access privileges could enable a person to delete or amend a basket
of transactions. This was not flagged in the body of the report by Deloitte as risk
area or a key matter and I did not identify the relevance of these rows when I
reviewed the report. However, I can now see that they contradicted James
Davidson's early statement that "There is no ability to delete or change records
branch creates in either old Horizon or Horizon online". I only appreciated later,
in early 2016, that Fujitsu may have the ability to edit and delete transaction data,
and I address this further below at §§299-302.
217. Having received the Desktop Report on 22 August 2014 I then spoke with Mark
Westbrook (Deloitte) on 3 November 2014 to discuss Deloitte’s findings. My
firm's telephone attendance note records that Mark Westbrook identified that: (i)
Balancing Transactions were “to be used in exceptional circumstances”; (ii) the
functionality of the tool and controls around its use were “best summed up in an
email from John Simkins (JS) at Fujitsu”; (iii) SPMs “probably did have visibility”
of injected Balancing Transactions, and indeed would “he imagine be initiated by
[them] as per the example found by Deloitte”; and (iv) establishing whether the
Balancing Transaction tool had been used in Old Horizon “would require a
detailed interrogation exercise on the old system which was not trivial.‘
165 WWBON0000908; POL00211255.
166 WBONO0000916.
Page 130 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
218. My firm's attendance note indicates that I felt that various follow-up enquiries
were needed, including obtaining a copy of John Simpkins’ email (which I did on
10 November 2014),1°” and asking Fujitsu to confirm that Balancing Transactions
were visible in branch accounts. I circulated an updated note on 10 November,
which would have reflected my understanding of the situation at the time, and
noted that this needed to be sent to Fujitsu for their review.’® I can see from the
draft note that I raised detailed further questions with Fujitsu regarding the effect
of Balancing Transactions and their visibility in branch accounts.
219. At this point, carriage of the note and responsibility for liaising with Fujitsu
appears to have passed to Mark Underwood (then an independent contractor
engaged to work on the Mediation Scheme).'® An update from him dated 20
November 2014 states:
“Whenever we have spoken to FJ about this issue, they seem puzzled
as to why we are so concerned citing ‘data integrity’ However I think we
are now of the opinion it is a semantics issue. By ‘data integrity’ FJ are,
I think, referring to ‘audit trail’ — in that, whatever is done leaves a clear
and identifiable audit trail behind it and thus — if there is no ‘remote
access car’ in the branch's data — it simply did not happen. This therefore
allows us to prove the negative.
On a call — FJ confirmed they already had downloaded all the branch
data available for the 150 scheme cases and performed searches for
any such ‘scars’.”"7°
220. This indicated that the branch data for each of the branches under consideration
in the Mediation Scheme (i.e. including those operating under Old Horizon,
187 WBON0000910; WBON0000911.
168 WWBON0000912; POL00212054. My covering email indicates that I also had a call with Fujitsu
around this time, though I have no recollection of this call and have not been able to identify
confirmed meeting arrangements in my email records.
169 WBONO0000914.
179 WBONO000479.
Page 131 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
subject to the limitation that the audit data only went back to around 2008) had
now been checked, and no evidence had been found of "remote access" e.g. an
injected transaction. This would seem to have been confirmation of what I had
previously understood to be overwhelmingly likely, namely that injected
transactions were extremely rare and not relevant to the cases within the
Mediation Scheme.
221. After this, Mark Underwood appears to have struggled to obtain useful further
information from Fujitsu.'’1 The version of the note that was eventually returned
on 10 December 2015'72 confirmed some important points (e.g. that “it is not
possible to edit existing transaction / basket data” and that Balancing
Transactions are “new transactions with unique jsn’s and identifiers’), but
otherwise failed to provide meaningful answers to a number of questions (e.g.
failing to answer a direct question about the visibility of Balancing Transactions
in branch accounts, and simply saying “See incident in March 2010” in answer to
the questions “When are SPMRs made aware that an injection is to occur?” and
“please describe the process and controls in place”).'’° It appears from my email
records that the note to Second Sight was never ultimately finalised.'74
Finalisation of Second Sight’s Part 2 Briefing Report
222. On7 April 2015, Second Sight notified POL that they had finalised their (updated)
Part 2 Briefing Report. They sought POL’s comments by close of business the
171 WBON0000917; WBON0000327.
172 Containing comments from James Davidson and Torstein Godeseth of Fujitsu.
173 WBON0000327.
174 POL00408247.
Page 132 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
following day on a substantially revised Section 14 of the report, which dealt with
remote access, together with two documents which they intended to cite. 17°
223. Both of the supporting documents referred to the Receipts and Payments
Mismatch bug, which had arisen in 2010 during the pilot of Horizon Online and
resulted in discrepancies being lost from the counter but retained on the back-
end system if a specific sequence of steps was followed by an SPM at the end
of a trading period. In particular:
223.1.The first document was an internal Fujitsu memo authored by Gareth
Jenkins, which set out the cause and effects of the bug, what information
needed (in his view) to be established before the matter was raised with
POL, and (subject to POL’s decision as to how to proceed) what steps
would be required to fix the data for each affected branch. It noted that
“[t]he data can be corrected by adjusting the appropriate Opening Figures
and BTS Data that relates to the current TP. This will result in the
Discrepancy needing to be processed when rolling over to the next TP”.
The memo recorded that “if we do amend the data to re-introduce the
Discrepancy, this will need to be carefully communicated to the Branches
to avoid questions about the system integrity” .‘7°
223.2.The second document was a note of a meeting between POL and Fujitsu
to decide how to proceed. One of the solutions discussed was to “Alter the
Horizon Branch figure at the counter to show the discrepancy. Fujitsu
would have to manually write an entry value to the local branch account’.
This would (as Gareth Jenkins’ note described) result in the branch having
175 POL00021845; POL00225912; POL00225913.
176 POL00225914.
Page 133 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
the discrepancy reintroduced, in effect realigning the counter and back-
end systems at the point of rolling over to the next trading period. It was
identified that this would have “significant data integrity concerns and
could lead to questions of ‘tampering’ with the branch system and could
generate questions around how the discrepancy was caused. This solution
could have moral implications of Post Office changing branch data without
informing the branch”.‘7”
224. The draft version of Section 14 of Second Sight's report observed that these
documents appeared to suggest that POL and/or Fujitsu did have the ability (at
least in 2010) to “alter” or “directly amend” branch data, and that this was
inconsistent with POL’s previous statements that no such facility existed. The
draft also referred to the Winn/Lusher email and commented that POL had not
explained whether or not it was accurate.
225. On 7 April 2015, POL put these matters to Fujitsu, who were asked to explain:
what was meant by the references to “adjusting” and “altering” data in the
documents; how such an alteration would be made; whether it would be visible
to the affected SPM; and what decision had in fact been taken in relation to the
Receipts and Payments Mismatch bug.'”® Fujitsu were also provided with a clean
version of the draft note on remote access which had last been updated in
December 2014 (which stated my understanding at that time that neither POL
nor Fujitsu could edit, manipulate or delete transactions).'79 Pete Newsome of
Fujitsu confirmed that the note appeared to correctly describe the process
177 POLO0225913.
178 WBONO000924.
179 POL00243542.
Page 134 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
referred to in the two documents supplied by Second Sight.'®° However, POL
recognised internally that it should not be complacent about the answers being
provided by Fujitsu,®’ and on further probing Pete Newsome confirmed:
“There is only one process Fujitsu_can use which is the insertion of
auditable _additional transactions described _in_the document so the
words below must have been a loose business description for a meeting
with nontechnical attendees.”'®*
226. I took the above as further confirmation from Fujitsu that it was not possible to
edit or delete transactions from branch accounts; the only route was for Fujitsu
to inject new ones. Fujitsu provided further information following a call on 8 April
2015 (which I cannot recall). In their follow-up email, they: (i) described the
process for making a Balancing Transaction in substantially the same terms as
John Simpkins’ earlier email to Deloitte; (ii) confirmed that “Any change would be
a new transaction in the audit log and can be identified under a separate
identifiable login in the branch audit record. All existing transactions are
unchanged’; (iii) reiterated that “this type of transaction will appear in the branch
printout’; and (iv) made the point that “[iJt is Post Office’s responsibility to explain
the need for the change and the change that took place with the Sub
Postmaster’ .1®°
227. On this basis, and with the deadline for commenting on the draft of Section 14
approaching, I (with others) prepared the following response to be sent by Patrick
Bourke to Second Sight:
189 WBONO000927.
181 WBONO000928.
182 WBONO000929.
183 POL00041040.
Page 135 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
“As we have always stated, Horizon does not have functionality that
allows Post Office or Fujitsu to edit or delete the transactions recorded
by branches ...
It has however always been possible for Post Office to correct errors in
and/or update a branch's accounts. Most commonly this is done by way
of a transaction correction however it could also be by way of a balancing
transaction or transaction acknowledgement ...
[I] can confirm that most of the branches affected by the Receipts /
Payments issue were resolved by Post Office writing off the
discrepancies ... In one branch, a balancing transaction was used to
correct the discrepancy in the branch's accounts (being ‘Solution 1’ in
the documents).
All of the above processes for correcting / updating a branch's accounts
have similar features. They are only used with a Subpostmaster's
consent, all of them involve inputting a new transaction into the branch's
records (not editing or removing any previous transactions) and all are
shown transparently in the branch transaction records available to
Subpostmasters (as well as in the master ARQ data).
Unfortunately, the language used in the documents produced by Post
Office / Fujitsu is colloquial shorthand that was only intended for internal
use by those who understood the Horizon system. I can understand why
these documents could be read to suggest that Post Office was ‘altering’
branch data but I hope the above explains why this is not the case.”'*4
228. On 9 and 10 April 2015, and after the final Part 2 Briefing Report had been
released, Fujitsu provided responses to certain outstanding questions that had
been raised with them, including how a Balancing Transaction would be identified
in branch accounts and the Audit data. In short, Mike Harvey stated that a
Balancing Transaction would appear in the branch accounts as a separate
transaction, but that “the Post Masters reporting does not go down to the level of
granularity to show that the transaction was an insertion [at the data centre].
However, the effect would be clearly visible ... [and] within the associated audit
log the use of the transaction correction tool would be clearly apparent and it
184 POL00041040; POL002260889. The final version sent is at: POLO0021785.
Page 136 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
would therefore be obvious that the transaction had not originated from the Post
Master or his/her team” .'°°
Review by criminal lawyers
229. At around this time, i.e. in early 2015, Cartwright King were considering the
Project Zebra Desktop Report and specifically whether the existence of the
Balancing Transaction functionality needed to be disclosed to convicted SPMs.
They provided POL with a note dated 27 March seeking further information about
the nature and use of the tool.'8° Reviewing my email records now, for the
purpose of preparing this statement, I believe that the original intent was for
Cartwright King’s questions to be forwarded to Fujitsu for comment,‘8” but that
POL’s correspondence with Fujitsu on the same subject shortly afterwards
rendered this unnecessary, and that I therefore collated the information which
had been provided by Fujitsu on the subject of Balancing Transactions in order
to answer Cartwright King’s queries.'® Self-evidently this was not advice (and I
was not asked to advise) on the substantive question of whether the Balancing
Transaction process should be disclosed to convicted SPMs, but rather it was an
exercise in gathering factual information to which I had ready access in order to
assist Cartwright King.
230. It is apparent from references within my emails that conferences were then held
with Brian Altman QC and Cartwright King to obtain their advice on how to
proceed. I understand from later emails that POL was advised to ascertain how
difficult and costly it would be to interrogate Old Horizon data to establish whether
185 WBONO000930.
186 WBONO0000922; POL00228075.
187 WBONO000340.
188 WBONO000931.
Page 137 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
the Balancing Transactions function had been used pre-2010, after which it could
be decided if this exercise needed to be commissioned in order for POL to comply
with its duties as prosecutor.'®° I therefore contacted Fujitsu for information on
this point and was provided with a note authored by Gareth Jenkins setting out
how such data could be accessed and examined.'° In short, Fujitsu advised that
“searching for BTs would in fact be an enormous task, taking several months of
work”, and my resulting instructions from POL were that it was “not prepared to
commission this exercise unless it is considered absolutely vital and there is no
credible alternative”.'°' I communicated this by email to Cartwright King on 15
July 2015, and to Brian Altman QC on 20 July 2015.19
231. Gavin Matthews and I attended a further conference with Brian Altman QC on 21
July 2015, who gave clear and unequivocal advice that “/t]here is currently no
need to give any further disclosure to SPMRs about BTs”. He considered that:
231.1.In relation to New Horizon, “the only BT was in a branch not touched by
any prosecution so there is no disclosure to give in this regard”.
231.2.In relation to Old Horizon, “POL does not have an obligation to go on a
fishing expedition, particularly one that would be extremely onerous and
costly’.
231.3.However, “[ijf POL knew that a prosecuted branch operating Old Horizon
had been subject to a BT, that specific fact may trigger a disclosure in that
specific case”.19°
189 See for example WBON0000944; WBONO0000946.
199 WBON0000942; POL00238791.
191 See for example WBON0000944.
192 WBON0000944; WBONO000946.
193 POL00021775.
Page 138 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
232. I conveyed this advice to Rodric Williams later the same day, together with the
fact that Cartwright King had been consulted but chose to defer to Brian Altman
QC.1% Rodric Williams subsequently asked me to obtain confirmation in writing
from Brian Altman QC that he agreed that my note accurately reflected his advice
(which I did).19° For the avoidance of doubt, and with reference to Q45.3 of the
Request, the principal reason for my involvement in these matters was that WBD
were Brian Altman QC’s instructing solicitors (see above at §§95-96) and it was
part of my role to obtain and act as a conduit for his substantive advice when
called upon to do so.
(vii) POL’s approach to deciding whether to mediate cases (Q38, Q47 to Q48)
233. The foregoing subsections set out my answers to the Inquiry’s questions about
how the process of investigating SPMs’ complaints unfolded. In this subsection I
turn to Q47 to Q48 of the Request, as well as aspects of Q38 (insofar as not
already answered above). These questions broadly concern my involvement in
“the decision-making process to determine whether to take cases to mediation”,
and (relatedly) my/WBD’s role in advising POL on the merits of individual cases,
including whether to mediate them.
234. There are three aspects which I address below. First, I set out the advice I gave
to POL concerning the role of the Working Group in deciding whether cases
should proceed to mediation, and the test which it ought to apply. Second, I
address the advice which POL received (much of which I did not give myself) to
guide its high-level decision-making in relation to which cases should proceed to
194 POL00021775.
195 POL00021777.
Page 139 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
mediation, and if so, what settlement criteria it should apply. Third, I summarise
WBD’s role (which I have already touched upon at §149.2(ii) above) in relation
to the application of these high-level criteria by POL to each case based on its
individual facts.
The role of the Working Group
235. The Working Group had to start considering the suitability of individual cases for
mediation around the middle of 2014, as this was when Second Sight started to
produce its first CRRs.1%
236. The Working Group meeting on 16 June 2014 involved the first contentious
discussion about whether to put a case through to mediation. This was case
M054. This case did not involve a convicted SPM, but it was a case where POL
(on WBD's advice) did not believe that a mediation was warranted. POL’s and
Second Sight's reports had both concluded that the applicant had admitted to
removing £9,500 out of an £11,900 shortfall from the branch immediately before
she was audited and was responsible for that part of the shortfall. POL believed
that the SPM was responsible for remaining shortfall (of c. £2,500) with Second
Sight unable to reach a conclusion on that point.9” On WBD's advice, POL did
not consider that it would be proportionate to mediate in respect of the remaining
(less than £2,500) shortfall, although it was prepared to enter into an informal
discussion with her about that aspect.'9® POL therefore proposed to (and did)
vote against mediation, whereas JFSA (supported by Second Sight) voted in
196 Cf. POL00026662; POL00026668.
197 POIR: WBON0000132; CRR: POL00306593.
198 WBONO000859.
Page 140 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
favour.'°° Sir Anthony Hooper was therefore called upon to exercise his casting
vote.
237. There was a discussion as to what test Sir Anthony (and indeed the Working
Group as a whole) should apply in deciding whether or not a case was “suitable”
for mediation in accordance with the terms of reference, and it was decided that
the test should be:
“On the assumption that both parties approach mediation in a genuine
attempt to reconcile their differences [is] it reasonably likely that the
parties will reach an agreed resolution of their issues”?2
238. This was formulated by Sir Anthony at the meeting, and I thought it was a
sensible test which accorded with the nature and objectives of mediation.
239. Sir Anthony retired to consider his decision on the suitability of case M054 for
mediation and on 24 June 2014 gave a written decision to the effect that: (i) POL
was reasonably entitled to conclude that a large part of the shortfall had been
removed from the branch by the applicant; (ii) POL would therefore be acting
reasonably in not agreeing to pay any part of that sum; and (iii) it was not
reasonably likely that mediation would lead to a resolution of the dispute.2”
240. JFSA was dissatisfied with this decision, and in July 2014 Alan Bates circulated
two separate proposals. First, that the test for “suitability’ should be reformulated
as “cases should proceed to mediation where mediation would allow the
Applicant an opportunity to express their concerns to Post Office”; and second,
199 POL00026664.
200 POL00026673.
201 WBONO0000864.
Page 141 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
that this test should be applied by Second Sight alone, without any oversight by
the Working Group.
241.1 (and those instructing me at POL) believed this represented the wrong
approach. It was a key part of the architecture of the scheme that the Working
Group, not Second Sight, should be responsible for considering the suitability of
individual cases for mediation. Further, we thought that Sir Anthony’s original
formulation of the test was the right one. There had never been any guarantee,
when the scheme was set up, that all cases would proceed to mediation and it
was clear that they would only do so if they were deemed “suitable” by the
Working Group. The test adopted on 16 June 2014 struck the right balance
between the interests of applicants and POL; to put any case through to
mediation regardless of merit on the basis that it would enable the applicant to
be heard would inevitably give many applicants false hope, at significant cost
and without any obvious benefit (particularly given that alternatives were
available to give applicants in this position a ‘voice’, such as a direct discussion
between them and POL). I prepared submissions to this effect on POL’s behalf
for Sir Anthony to consider.2°>
242. On 26 August 2014, Sir Anthony determined that the Working Group was the
proper entity to make decisions on suitability, but that the test to be applied was
that proposed by JFSA.2 In reaching this decision, Sir Anthony considered that
the guide to the Mediation Scheme gave the impression that most cases would
202 Concerning the first proposal, see: POL00026671; WBONO0000876 and attachments:
WBON0000877, POL00206822, and POL00206823. JFSA’s proposed new suitability test is
captured here: WBON0000874. Concerning the second proposal, see POL00026672;
POL00207229.
203 WWBON0000885; WBON0000886; and POL00207393; POL00207394.
204 WBON0000893; POL00210134.
Page 142 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
be mediated, because it had said “/f your case is suitable and you provide
accurate, detailed information to Second Sight, then this [mediation] is likely in
most circumstances”.
243. I did not consider that this was the right reading of the guide, which (to my mind)
stated that mediation would be likely “in most circumstances” if the case was
deemed suitable and the applicant had provided sufficient information.2°>
Although POL had to do its best to work to Sir Anthony’s decision in good faith,
it posed real difficulties as it effectively meant that cases would be routinely
approved for mediation by the Working Group, even though a case may have
little hope of reaching a settlement or the SPM had an extant criminal conviction
(and as explained below, POL had received firm advice from its criminal lawyers
that it could not countenance mediating such cases). Whilst Sir Anthony's
decision did not mean that POL would be forced to mediate cases when it
deemed this inappropriate (the Working Group having no power to compel this),
it did mean that POL was put on a course where it would have to refuse to
mediate cases despite that being the Working Group's recommendation. This
situation meant that the Working Group was no longer collaboratively filtering out
unsuitable cases (or was much less likely to do so), which was important if
disappointed applicants were to understand and accept the decision that their
case should not be mediated?
244. POL was sufficiently concerned about the position in which it found itself that it
obtained specialist public law advice (from DAC Beachcroft) about the risk of a
205 WWBONO000895.
206 See for example the considerations outlined in POL’s internal briefing notes for the Working
Group meetings on 2 and 17 October 2014: POL00210056; POL00211024.
Page 143 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
judicial review claim being brought if it declined to mediate cases which the
Working Group had approved.?°” This advice confirmed that the public law risk
was low.
Advice on mediating criminal cases
245. As I mentioned earlier (§136), POL’s criminal lawyers had in September 2013
expressed misgivings about the decision to allow SPMs with a relevant criminal
conviction (that is, for theft, fraud or false accounting) to apply to the Mediation
Scheme. By that stage, though, the decision to entertain applications from such
SPMs had been taken. For my part, I could see that there might be some practical
value in allowing these cases to be considered as part of the investigative phase
of the scheme, since this might turn up relevant information or even simply
enable the applicant to better understand the decisions POL had taken in their
case.
246. However, once the Working Group started to make decisions on suitability in
summer 2014, POL had to determine how it should approach the question of
mediating such cases. It sought advice from Cartwright King, who on 8 July 2014
strongly advised against mediating such cases.?° I understood that they were
concerned that the mere fact of entering into mediation with a convicted SPM
could be seen to cast doubt on a conviction which might be otherwise safe. This
was discussed at a Working Group meeting and JFSA objected in strong terms
to POL’s suggestion that it could not mediate with convicted SPMs.2°
207 WBON0000900; WBON0000902.
208 WWBON0000867; POL00305248; see also WBONO000869.
209 WBONO000870.
Page 144 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
247. Sir Anthony Hooper developed a proposal whereby, instead of mediating with a
view to settlement, the parties would mediate with a view to POL deciding
whether it would, or would not, support an appeal against the SPM's conviction.
A conference was sought with Brian Altman QC,?"° and on 15 September 2014
he provided a formal written advice on the approach POL should take in such
cases.?"t In short, he advised in clear terms that Sir Anthony’s proposal was
unworkable since POL could not be seen to take a position on the safety of a
conviction before the matter had been considered by the Court of Appeal.
Further, he advised strongly against POL mediating in false accounting cases
with a view to identifying the cause of the loss that led to the rendering of the
false accounts, which he saw as fraught with difficulty. The only thing POL could
do in relation to criminal cases, in his view, was to disclose any material identified
during the investigative phase which cast doubt on the safety of the conviction,
but it should not countenance mediating such cases. POL accepted this advice
though it led to serious difficulties within the Working Group for the reasons I
have outlined above.?'?
Linklaters advice and generic settlement criteria
248. In addition to the advice obtained from the criminal lawyers, POL obtained advice
from Linklaters about its potential contractual liability based on the types of
complaints that were being raised by SPMs through the Mediation Scheme. That
advice was provided on 20 March 2014, and, in relation to complaints alleging
that the Horizon system was responsible for the SPM’s shortfalls, it stated that
210 WBON0000871; WBON0000406.
211 WBON0000900; POL00214992.
212 See, for example, in relation to case M030: WBON0000906; WBON0000903; WBON0000905;
WBON0000907; WBON0000407.
Page 145 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
POL should not consider settling in the absence of clear proof that Horizon was
not operating as it should.2'? It did however outline certain limited categories of
case where POL might consider making some concessions. I (together with WBD
colleagues) had earlier prepared a draft settlement policy outlining generic
settlement parameters in the types of cases being considered under the
scheme,2"4 and on 26 March 2014 I provided a supplemental advice giving
guidance on valuing claims for harm caused to the value of an SPM'’s retail
business following the allegedly wrongful summary termination of their
contract.2"5
Case-specific decisions
249. I have already explained above (at §149.2(ii)) that once Second Sight produced
its CRR, the WBD team would review the case and produce a short written note
of advice setting out whether a case should be mediated and if so what the
settlement parameters would be. This would inform POL’s decision as to how to
vote at Working Group meetings, as well as POL’s settlement parameters for that
case if it proceeded to mediation.
250. These advice notes were based on, and applied (i) Brian Altman QC’s advice in
relation to mediating conviction cases; (ii) Linklaters’ advice; and (iii) the
guidance on settlement criteria provided by WBD, read subject to and in light of
the other advices. I exhibit hereto a selection of these advice notes to illustrate
their form and content.?16
213 POL00202008.
214 POL00202008.
218 WBON0001707; POL00278283.
216 I have been advised that POL consider the documents I had referenced here to be privileged
and that it is not willing for such privilege to be waived.
Page 146 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(vi) Closure of the Working Group (Q35.6 and Q50 to Q52)
JFSA’s objections
251. With reference to Q50, the nature of the concerns JFSA expressed in its letter to
Sir Anthony Hooper dated 10 November 2014 (i.e. the letter that was the subject
of the discussion at the Working Group meeting on 14 November 2014, the
minutes of which are POL00043630) were in brief summary as follows:2"7
251.1.JFSA maintained that it was not the Working Group’s place to decide
whether cases were unsuitable for mediation; they believed that SPMs had
been promised, when the Mediation Scheme was established, that a//
cases would proceed to mediation as long as the applicant provided
sufficient information about their complaint at the outset. Their belief was
that the only issue for the Working Group was whether or not an applicant
had provided enough information to be admitted onto the Scheme in the
first place.
251.2.They were particularly aggrieved by POL’s position that it could not
generally mediate with SPMs who had extant criminal convictions. They
were also upset by the fact that POL considered that mediation was not
the appropriate route in cases that lacked merit, but the real crux of the
issue (in my view) was the impasse between POL and JFSA on criminal
cases.
217 POL00216273.
Page 147 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
251.3.JFSA also expressed the view that where mediations were entered into,
POL’s conduct was not sufficiently conciliatory and that mediations were
not leading to resolutions.
252. These factors — which were interconnected — led the JFSA to the belief that the
Mediation Scheme was a “sham” which POL was not engaging with in good faith.
They ultimately disengaged from the Working Group — see the minutes of the 8
December 2014 meeting at POL000043631 — which caused decisions on
borderline or contentious cases to not be made or to be deferred.
253. I have no recollection of the discussion on 14 November 2014, referred to in
POL00043630, but I recall that in general terms, I did not think that JFSA’s
criticisms were fair or well-founded:
253.1.To my mind, it had always been clear that some cases might not be
mediated (and that these were not confined to cases where the applicant
failed to provide enough detail to progress an investigation into their
case).2"8 As I have explained above, §§241-243, the guide to Mediation
Scheme said, at p.8:
“Will my case definitely be referred to mediation?
If_your_case_is suitable _and you provide accurate, detailed
information to Second Sight, then this is likely in most
circumstances.
However, the Working Group may consider that some cases are
not suitable for mediation. For example, if there is insufficient
information about a case or the case_is_not_one requiring
resolution.
218 See an email I sent setting out my thoughts at the time: WBON0000915.
Page 148 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Also, once Second Sight has submitted its findings, Post Office
may contact you to discuss your case and to seek a resolution
without needing to attend mediation.
If your case is not referred to mediation, then you may still pursue
other methods of resolution such as by bringing a claim through
the Courts.’219
253.2.1 cannot recall whether POL was explicit with JFSA, when setting up the
Scheme in Autumn 2013, that this might mean that convicted SPMs’ cases
might not be mediated. I recall that there was a discussion around
available remedies, i.e. POL made clear that the Mediation Scheme could
not result in any convictions being overturned. To the extent that POL did
not go further and explain that criminal cases would not be mediated at all,
I do not believe I saw any evidence that this was the product of bad faith
on POL’s part. Rather, (i) it had not yet obtained formal advice from its
criminal lawyers on this point, and (ii) in any event, allowing such cases
into the Scheme might produce benefits because the investigation might
surface documents or information that could assist in an appeal against
conviction.
253.3.It is worth adding that it is something of an oversimplification to say that
POL's position was that it could not mediate with convicted SPMs at all.
There were a couple of conviction cases which POL did agree to mediate,
because the factual basis of the SPM’s complaint was not directly related
to the conviction.22° This was reflected in an internal WBD briefing note
which gave guidance to the team on the conduct of mediations, and
218 WBONO0000805.
220 See: POL00218712, p.3.
Page 149 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
explained that POL’s position was that it “will not mediate any cases
involving a criminal offence save in exceptional circumstance” .7*'
I was not of the view that POL were refusing to engage properly in the mediation
process. From my perspective, WBD’s instructions (as reflected in the
abovementioned briefing note) were to engage in mediations in good faith with a
view to achieving resolution, or if that was not possible, to “/t/ry to bring closure”
or at least “make sure the Applicant has had a fair chance to put their position to
POL and to ensure that POL has constructively responded”. Unfortunately, I did
get the impression that there was something of a disconnect between JFSA's
and SPMs'’ expectations, and what POL were willing to offer. POL were not
prepared to offer financial settlements in cases where it considered it had little
risk of legal liability (since ADR processes such as mediation were, at bottom, a
way of achieving an out of court settlement of a legal dispute). By contrast, JFSA
and SPMs seemed to think that the fact that a mediation was happening meant
that a financial offer would definitely be made. I can’t remember whether I had
already formed this view at the time of JFSA’s letter as, in my recollection, only a
few cases had proceeded to mediation by this stage. But it was certainly a view
I formed as more mediations took place, and the review of the mediation process
prepared by CEDRat the end of the scheme confirmed their view that applicants
often approached mediations as though they were actually in a compensation
scheme.?22
Decision to close the Working Group
221 POL00407979.
222 POL00232900.
Page 150 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
254. Q51 asks me to set out the nature and extent of my involvementin POL’s decision
to terminate the Working Group and mediate all cases, save those involving an
SPM with a relevant criminal conviction. This was a policy decision; I did not
advise on that aspect of the decision and had no involvement in POL’s wider
political and strategic decision-making in relation to the Scheme (for example,
whether to continue it at all, whether to close it but offer a financial settlement to
some or all applicants in lieu of mediation, and so on). Searches of my email
records indicate that the Project Sparrow Subcommittee made a
recommendation that was then adopted by the Board.?25
255. To the best of my knowledge, my involvement in these matters consisted of:
255.1.Providing brief advice on the proposal to terminate the Working Group,
prior to the Project Sparrow Subcommittee’s consideration of that issue in
early February 2015 (this is the advice shown in POL00021908);
255.2.Providing POL with formal written advice on the “manner of
implementation and consequences’ of its related proposal to terminate
Second Sight’s contract for services, also in early February 2015.24 I
exhibit that advice hereto.2#
256. In relation to my advice on the decision to terminate the Working Group, and with
reference to Q52.1 of the Request, I confirm consider the advice contained in my
email to Belinda Crowe dated 9 February 2015 (POL00021908) to have been
legal advice. I was instructed to comment on the proposal contained in the
223 WWBON0000408; WBONO0000409.
224 POL00021728. From my emails, Victoria Brooks picked up the drafting of the advice itself,
under my supervision: WBON0000921.
225 POL00221480.
Page 151 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Project Sparrow Subcommittee paper attached to that email,?2° and in so doing
would have been concerned with whether the proposal would meet POL’s stated
objective, and what the attendant legal risks were. The objective was that
specified at paragraph 2.12 of the paper, namely that POL should continue to
“meet [its] obligations to applicants but also seeking to regain control of the
process and bring it to a conclusion as soon as possible”.
257. As my email to Belinda Crowe at POL00021908 shows, I agreed with the
proposal to terminate the Working Group as an effective means of achieving this
objective, because:
257.1.First, against the background set out above, the Working Group had
become fractious and largely inoperable, with JFSA refusing to participate
in discussions or votes. Further, the procedures for progressing cases
through the Scheme were by now well-established — thanks to the efforts
of the Working Group — meaning that this aspect of its role had significantly
reduce by February 2015. Taken together, these factors meant that in my
view, the Working Group “offer[ed] no real value” any longer (cf. Q52.2).
257.2.Second, closing the Working Group by taking a decision to mediate all
cases (save for non-criminal cases) would accelerate the conclusion of the
Scheme by eliminating the intermediate step of the Working Group
considering and voting on those cases. Incidentally, and with reference to
Q35.6 of the Request, I thought that the Working Group had fulfilled its
purpose at this stage (notwithstanding the fact that it largely broke down
towards the end) because of this combination of factors. Its twin purposes
226 POL00221561.
Page 152 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
were (i) to establish the process for progressing cases through the
Scheme to the point of deciding on their suitability for mediation, and (ii)
considering their suitability for mediation. As I have explained above, it had
achieved (i), and the proposal in the Subcommittee paper would relieve
the need for (ii).
257.3.Third, the fact that the breakdown of the Working Group was “the source
of much criticism” was a material consideration (cf. Q52.2). On the one
hand, it was leading to POL being perceived as the obstructive party, and
as failing in its obligations to Scheme applicants. This was not just a matter
of reputational concern; it undermined confidence in the Scheme itself and
therefore reduced the likelihood of mediations achieving resolution. On the
other, it was exacerbating delays which again was contrary to the
objectives of the Scheme.
258. As to Q52.3, again, I was commenting on the efficacy of the proposal in the
Project Sparrow Subcommittee paper. That proposal was for POL to “meet its
commitment to any applicant wishing to avail themselves of a review by Second
Sight of their case by providing the necessary funding to do so on an individual
basis” (p.4). As my email to Belinda Crowe makes clear, my concern was that
“[i]f SS are independently contracted _fi.e. paid] by Applicants” then their
independence would be lost. I anticipated that if Second Sight were dependent
on applicants for their funding, they would increasingly become advocates for
those applicants (and that this would likely involve them straying further beyond
the bounds of their expertise even than they had done to date).
259. As to Q52.4, Second Sight had in their possession a large number of documents
about applicants containing private and sensitive data — including criminal
Page 153 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
convictions data. As my email to Belinda Crowe shows, I was concerned that
Second Sight might not hold that data securely or might release it to third parties.
This could have serious implications for POL (as well as for the SPMs concerned)
because (i) POL was the controller of that data and therefore ultimately
responsible for it, and (ii) POL owed Mediation Scheme applicants strict
obligations of confidentiality.
H. KNOWLEDGE OF BUGS, ERRORS, DEFECTS, AND REMOTE ACCESS,
2013 to 2015 (Q6 to Q9)
260. The preceding sections set out my answers to the Inquiry’s questions in respect
of the period from when I was first instructed in relation to the Horizon-related
matters in April 2013, to mid-2015 when my work on the Mediation Scheme was
substantially complete. In the course of answering those questions I have where
relevant addressed my knowledge of bugs, errors and defects within Horizon,
and of remote access to the Horizon system, at the time in question.
261. In this section I summarise how my knowledge and understanding of these
matters evolved during this period, as well as what I then thought of the
‘robustness’ of Old Horizon and Horizon Online, in order to answer to Q6 to Q9
of the Request. Given the extent of my involvement over the period in question
(which continued for several more years thereafter), and the lapse in time, I
cannot recall every bug, suspected bug, or remote access issue that was
addressed. The below paragraphs are therefore only a summary and are not
exhaustive of all my knowledge, but I have done my best to capture all the
Page 154 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
material points I can recall and which informed my understanding of the likely
robustness of the Horizon system.
262. It is important to highlight a few matters at the outset which are material to my
state of knowledge during this period (and indeed more generally). I have
touched on some of these matters above, but they bear repeating in order to
contextualise what follows:
262.1.First, I was at all times acting as an external solicitor. Accordingly, my
knowledge derived entirely from the information and instructions which
POL provided, and information obtained from Fujitsu (and to a lesser
extent, Deloitte) where relevant.
262.2.Second, I was reliant on subject matter experts to understand the
technical detail as well as the practical significance of the factual
information I received in relation to the Horizon system. In this regard, it
was apparent to me from early on that the relevant expertise was drawn
almost entirely from Fujitsu. Although there were many people within POL
who understood Horizon well from a user’s perspective (e.g. trainers,
support teams, branch accounting people, etc), including the POL IT team,
I did not come across anyone at POL who had a detailed understanding
of the technical workings of Horizon. The main consequence of this, as I
understood at the time, was that POL was not in a position to itself
interrogate (still less verify) the factual information and technical advice
which emanated from Fujitsu. Whilst I and others at WBD could (and did)
seek to ask probing questions — as employees of POL also often did, from
my observations — this dynamic served to accentuate POL's dependence
on Fujitsu as a reliable source of information.
Page 155 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
262.3.Third, as I have explained above, my starting point when I was first
instructed was that I had little if any familiarity with the Horizon system,
and from there I learned about how the system worked progressively over
time. I was not, for example, given a formal or detailed briefing on the
functionalities or technical operation of the Horizon system (though as
explained above, I recall the basic differences between Old Horizon and
Horizon Online being explained to me at some point, as well as the role
and significance of the Core Audit Process). This is not intended as a
criticism, as it would have been a massive and complex endeavour and
not necessary or proportionate to my role as an external solicitor dealing
with the instructions I have described above. But it did, on reflection, mean
that it was not always straightforward to identify errors or subtle
inconsistencies in the instructions and information I received, given (for
example) the volume of the information passed to me over a long period
and the highly technical language in which much of that information was
expressed.
(i) Bugs, errors and defects (Q6)
263. I understand “bugs, errors or defects in Horizon” (which I shall generally refer to
as "bugs") to mean faults in the software which caused or contributed to
shortfalls or gains in branch accounts. I do not include here faults in Horizon that
did not have an impact on the accuracy of branch accounts because the centre
of all these matters is shortfalls in branches and whether SPMs or POL were
responsible for them.
Page 156 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
264. It is important to bear in mind that during my involvement there were different
investigations by different individuals and/or entities into various alleged issues
with Horizon. Some of these were directly linked to Horizon (e.g. a known bug),
but the majority were circumstantially suggesting a potential problem with
Horizon (e.g. an SPM would say that their accounts were wrong because they
suspected a problem with Horizon, but could not be more specific than that), or
concerned matters ancillary to the Horizon software (e.g. problems with printers,
or issues with the training and support services provided by POL).
265. In relation to the large majority of the accounting discrepancies which were
investigated over the course of my engagement, it was believed by POL that
those issues were not in fact caused by bugs, but rather it was able to present a
clear, or at least credible, alternative explanation for the issue. By way of
example, the transaction reversal process (which was the subject of Spot Review
1) was not considered by Second Sight to be an instance of a fault in Horizon.
Rather, Second Sight concluded in relation to this process that “the Horizon
system did operate in accordance with its design”,?2’ albeit that that design was
complex and difficult for SPMs to satisfactorily operate. I referred to the
transaction reversal process above at §§82-84, §171 and §174.
266. The only confirmed bugs of which I was aware during the period I am currently
considering, namely April 2013 to mid-2015, were:
266.1.The Callendar Square bug;
266.2.The Receipts and Payments Mismatch bug; and
227 Interim Report, §1.13: POL00099063.
Page 157 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
266.3.The Suspense Account bug.
267. I learned about each of these early on in my engagement. Indeed, I would have
had some awareness of the Callendar Square bug prior to my engagement
because it was referenced by the High Court in the Castleton judgment; as
mentioned above (§24) when I read this case early in my career. However, at
that stage I would have been I focused on the legal conclusion the Court reached
that SPMs could be pursued in debt on the basis that they had submitted a
‘settled account’ to POL, rather than on its treatment of the Callendar Square
bug.
268. I became aware of the Suspense Account bug when I was asked by Rodric
Williams to prepare a summary of the effect of the Suspense Account bug and
the steps POL had taken in relation to it.2?8 From my emails, it appears that I also
assisted in preparing letters to SPMs who had been affected by the bug.”
269. In relation to the Receipts and Payments Mismatch bug, the earliest reference I
have identified in my email records is my email to Rodric Williams of 1 July 2013
giving an overview of the Misra case (see above, §57).° From that email it is
evident that I was aware of the bug at the time of writing; I cannot say exactly
when I learned of it but I am confident that it would have been not long before
this, and in the same context i.e. assisting POL to prepare for receipt of the
Interim Report. It would not have been via a Spot Review, as the Spot Review
process did not reveal any bugs. It was not part of my instructions from POL
228 POL00407496.
228 POL00407493; POL00407494.
230 WBONO0000746.
Page 158 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(which at this time were primarily focused on the Spot Reviews) to investigate
this bug any further.
270. I did not learn of any other bugs known to have affected the operation of Horizon
during the period covered by the sections above. To my knowledge, none were
identified in the course of the rest of Second Sight’s investigation or during the
Mediation Scheme. As explained above at §§98 ff, the Horizon Regular Call was
established as a forum for departments within POL to share and investigate
information about issues reported with Horizon, and so far as I was aware this
forum did not confirm any new bugs in the system that were known to impact on
branch accounts (although I did not personally attend any calls after February
2014).
(ii) Remote access and suspicions about Fujitsu (Q7 to Q8)
271. I understand ‘remote access’ to relate, in particular, to the ability that Fujitsu
employees had to add to, edit, delete or otherwise influence transaction and
branch accounting data, including their ability to do so without the knowledge or
consent of the SPM concerned. It is my awareness of those capabilities that I
focus on here, rather than (for example): any read-only capabilities that POL and
Fujitsu had; POL’s ability to submit Transaction Corrections and Transaction
Acknowledgments to SPMs; or the ability of certain POL employees to conduct
transactions whilst physically in branch (thus, not remotely) using a “Global User”
account.
272. In the period from April 2013 to mid-2015, in summary, there were two main
issues around which my knowledge of the question of remote access coalesced:
Page 159 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
272.1.First, the allegation by Michael Rudkin that he had witnessed a POL
employee in Fujitsu’s Bracknell office demonstrate an ability to pass
transactions directly into Horizon in August 2008, which was the subject of
Spot Review 5 (and later, case M051 within the Mediation Scheme). I have
explained WBD’s involvement in this at §§38-50 above; suffice it for
present purposes to say that my instructions, and my belief, were that Mr
Rudkin was mistaken and that my understanding was that there was no
remote access functionality of the kind he had described.
272.2.Second, Fujitsu's ability to inject transactions into branch accounts (but
not, as I was informed at the time, to delete or edit existing data). At §§202-
232 above, I have described in detail the extent of my awareness of these
investigations, and the impact that they had on my knowledge.
273. By mid-2015 I was under the impression that:
273.1.Robust controls existed to ensure that data was accurately recorded and
stored in the Core Audit Log.
273.2.Neither POL not Fujitsu had any ability to edit or delete transaction data.
273.3.The only method by which POL could push new transactions into branch
accounts from outside the branch itself was via the Transaction Correction
or Transaction Acknowledgment process.
273.4.1 was aware that Fujitsu could inject a transaction without the consent of
the SPM concerned, but I was reassured that this was not unconventional
or surprising in the context of a system like Horizon, and that written
protocols were in place to ensure that the use of the tool was strictly
controlled (for example, POL’s permission was required, it was required to
Page 160 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
be carried out under supervision of a witness, and there would be a clear
audit trail associated with any use of the tool).
273.5.There was some uncertainty as to how obvious it would be to an SPM if a
transaction was injected into their accounts, but I was reassured that: it
would be visible (in the sense of appearing as a separate transaction); it
was clearly marked as an injected transaction in the master audit data;
and injected transactions were a measure of last resort that would only be
used in exceptionally rare circumstances, such that my understanding was
that an SPM would be proactively consulted first.
273.6.The exceptional nature of injected transactions was demonstrated by the
fact that only one Balancing Transaction had been entered in the lifetime
of Horizon Online (i.e. once in approximately 4 years). Fujitsu was
absolutely clear on this point, and Deloitte had likewise found only one use
of a Balancing Transaction.
273.7.Fujitsu had conducted a review of the audit data for the branches of
applicants to the Mediation Scheme and confirmed there were no signs of
remote access (subject to the limitation that the audit data only went back
to around 2008).
274. Against this background, I did not at this stage suspect that Fujitsu’s ability to
manipulate transaction data was more extensive than set out above, or that they
had made more extensive use of this capability than they had indicated. In April
2015, in response to a direct question about references in a Fujitsu document to
‘adjusting’ and ‘amending’ branch accounting data, a senior Fujitsu employee
had said in terms that “/t/here is only one process Fujitsu can use which is the
Page 161 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
insertion of auditable additional transactions’.2°' Moreover, my experience of
Fujitsu in general was that they appeared to investigate POL’s queries about
Horizon thoroughly and in detail. Whilst there were some difficulties in getting
clear and timely answers from them about the Balancing Transaction tool (albeit
that I was not involved in all of those interactions), I had the impression that this
was largely because they thought of the issue as something of a red herring, not
because they had anything to hide.
(iii) Views on ‘robustness’ (Q9)
275. There are different ways of understanding and expressing what is meant by the
‘robustness’ of the Horizon system. In considering this issue during the period
from April 2013 to mid-2015, my focus was on whether SPMs were being held
responsible for discrepancies and shortfalls in branch accounts due to problems
in the Horizon software. Based on my involvement to this point, I did not believe
that this was a likely explanation, and accordingly I believed Horizon to be a
robust system. I identify the following as the key reasons for this.
276. First, these were my instructions from POL. That view was also supported by
Fujitsu. I appreciated that I could not simply rely on Fujitsu, but they knew the
system best and were a highly-respected IT company.
277. Second, the number of complaints from SPMs about Horizon was small. POL
had around 11,000 branches and thousands of SPMs at any one time. Added to
that would be hundreds or thousands of leavers and joiners over the course of
the 15 or so years that Horizon had (by 2015) been in place. Despite that, only
231 WBON0000929; and see above, §225.
Page 162 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
around 50 SPMs joined the Spot Review process, and around 150 applied to join
the Mediation Scheme. Against this context, it did not seem to me that a
significant proportion of SPMs considered that Horizon was unreliable. To the
contrary, the fact that only a small proportion of SPMs had come forward with
any complaints at all suggested to me that, for the vast majority of the time,
Horizon was working well in branches.
278. Third, Second Sight concluded in their July 2013 Interim Report that "We have
so far found no evidence of system wide (systemic) problems with the Horizon
software". They did not resile from this central conclusion in any of their later
reports and I do not recall any case in the Mediation Scheme where they
concluded that a shortfall was caused by a bug or other malfunction in Horizon.
Whilst I had reservations about the quality of some of Second Sight’s work, after
nearly three years of working consistently on Horizon and studying individual
SPMs’ complaints, my expectation was that they would have found some
evidence of system-wide software problems causing accounting discrepancies,
if any existed.
279. Fourth, a large proportion of the complaints which I had seen during both the
Spot Review process and the Mediation Scheme were not focused on Horizon,
but rather were about POL's training and support processes, or the user-
friendliness of specific accounting processes. Whilst I had sympathy with some
of the criticisms that SPMs and Second Sight made about the usability of the
system, these matters were not fundamentally about bugs in Horizon and so did
not, in my view, call into question that Horizon was robust.
280. Fifth, to the extent that complaints made during the Spot Review process and
Mediation Scheme were directly concerned with the Horizon system, my overall
Page 163 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
impression was that POL and Fujitsu took them seriously, investigated them
conscientiously, and were able to produce credible answers explaining why the
issues raised were not the result of bugs which indicated that Horizon was not
operating as intended.
281. Sixth, although by Summer 2013 three bugs in Horizon had been identified of
which I was aware, to my mind these did not lead to a conclusion that the system
was not robust having regard to the number of branches affected, the fact that
the errors had been corrected, and the overall scale of the Horizon system (which
was such that I thought it unsurprising that bugs would occasionally occur).
Indeed, Second Sight was aware of these bugs at the time of producing their
Interim Report, yet still considered that there was no evidence of system-wide
defects. The fact that no further bugs were (to my knowledge) identified in the
two years thereafter reinforced my impression that Horizon did not suffer
systemically from bugs so as to be the root cause of the discrepancies and
shortfalls which were the subject of SPMs’ complaints in the Spot Reviews and
during the Mediation Scheme.
282. I only add that, whilst I note that the Request asks about the extent to which I
‘believed’ in the robustness of the Horizon system (and I have endeavoured to
answer that point above), it should be borne in mind that, at the time of my
engagement, it was incumbent on me to consider whether, overall, POL had a
properly arguable case that Horizon was a robust system. As a solicitor acting
for a corporate client, I had a duty to follow my client's instructions and act in its
best interests within the scope of my retainer and professional duties. The work
which I carried out for POL throughout the period considered above was
necessarily guided by these considerations. Up to mid-2015 my role had
Page 164 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
involved, in summary, looking at specific points in the Spot Reviews, supporting
the Mediation Scheme which was examining the events in individual SPM cases
to see if Horizon was at fault in those cases, and assisting POL in responding to
ad hoc technical questions raised by Second Sight as part of their ongoing
investigations. Beyond this, I was not instructed to investigate bugs in Horizon.
In any event, I believed that these activities by POL were a reasonable way to
surface the existence of bugs and, ultimately, to identify whether problems in
Horizon were the cause of SPMs being held liable for shortfalls.
1. ADVICE ON POL’S EXTERNAL COMMUNCATIONS STRATEGY (Q53 to Q54)
283. My involvement with POL’s external communications strategy was ad-hoc. On
occasion, POL would ask me for specific pieces of information that would feed
into communications statements, or to review documents or proposed
statements for accuracy and to ensure they did not create legal risk from a civil
litigation perspective. I would also flag issues that I believed could cause
reputational harm. POL was a brand-focused organisation, however I did not
consider this to be unusual for a retail business like POL and overall my approach
to highlighting reputational risk was, I believe, typical of a commercial solicitor
advising a corporate retail client.
284. I did not set the communications strategy; this was beyond the scope of my
instructions and expertise. POL had an in-house communications team and
engaged specialist solicitors (generally CMS Cameron McKenna) to advise it on
media-related issues.
Page 165 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
285. Towards the end of 2014 and during 2015, POL publicly defended its position
that Horizon worked. This was in connection with coverage by the media (and
especially the BBC) of, among other matters, SPMs’ claims that errors in the
Horizon system had been the cause of shortfalls they suffered, and that they had
been wrongly and unfairly prosecuted for theft and/or false accounting as a result
of these shortfalls. It appeared to me to be reasonable that POL should take
steps to defend its position because, according to my instructions and based on
the information of which I was aware at the time, Horizon was highly unlikely to
be the cause of the SPMs’ shortfalls. I have been asked to comment on an email
which I sent to Rodric Williams on 18 August 2015 (POL00021865). My vague
recollection of this line of correspondence is that the BBC Panorama
documentary had aired the day before, and POL wanted to understand to what
extent (i) it was bound by any legal obligations of confidentiality to SPMs and
others when responding to the allegations made in that documentary, and (ii)
those who had commented in the documentary, and particularly Second Sight,
had breached any duties of confidence they owed POL. Second Sight’s terms of
engagement in relation to the Mediation Scheme contained provisions relating to
confidentiality, and, as I explained above (§142), Working Group discussions,
investigations into SPMs’ cases during the scheme, and mediations between
POL and SPMs were all intended to be confidential and subject to without
prejudice privilege. I was therefore the natural person to advise in the first
instance given my proximity to the scheme.
286. My advice was that POL was permitted under the confidentiality obligations to
comment on individual SPM cases, but my view was that POL should issue
nothing more than a denial of the allegations and wait for the CCRC investigation
Page 166 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
to conclude. This was because POL's "arguments are technical, rely on the
intricacies of the prosecution process and are based on a range of evidence
rather than a single smoking gun. The man in the street will simply say: “where
is the evidence of the SPMR putting the money in her pocket?”. I don’t believe
we can win this battle in the media”, I felt that there was a general public
perception that in order for an SPM to be morally culpable for false accounting
they would also need to have physically taken money out of a branch, and thus
that there would have been some evidence (e.g. CCTV or personal bank account
statements) of them taking the money. In many of the cases I was aware of, it
was not this straightforward. The evidence of false accounting was often in the
patterns of accounting that had the effect of obscuring the true accounting
position. I thought it would be hard to convey this complexity in a soundbite to
the media.
287. Thus, the phrase “intricacies of the prosecution process” was intended as a
shorthand for my understanding of what amounted to false accounting versus
the public perception of what that might be. I also had in mind a concept which I
had been told by Cartwright King was the basis for the theft charge in some of
the criminal cases, namely that, if there was direct evidence of false accounting
and no other explanation was provided by the SPM for the missing money, it
could be inferred that there had been fraud or theft of the money by the SPM. I
always thought that this argument involved an intricate analysis because it meant
that SPMs could be prosecuted for fraud or theft even though there was no direct
evidence of this (e.g. CCTV footage on money being moved).
288. The “range of evidence” to which I was intending to refer was the SPMs’
accounts, including their falsification of the same, plus sometimes I had seen
Page 167 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
cases which included supporting paperwork from within a branch, interviews with
customers and assistants, and/or admissions by the SPM. I was intending to
convey the point that there may not be a single, easily explicable piece of
evidence that was conclusive of an offence.
289. As I explain above, I did not believe that POL could “win this battle in the media”
because the public perception was that theft would have involved something like
a physical act of removing money from the counter, yet the legal position and
evidence was (as I understood it) more complex than this. I was concerned that
POL appeared to be on the back foot in this respect, because I felt that if the
media attention on this issue continued it could result in someone being prepared
to fund litigation against POL, as eventually happened.
J. THE SWIFT REVIEW (Q55 to Q57)
(i) Involvement prior to the group litigation
290. The “Swift Review” was commissioned by POL in-house and was conducted on
behalf of POL’s then chair, Tim Parker. I was aware of POL's intention to
commission a review and briefly discussed it with Patrick Bourke and Rodric
Williams at POL; I recall that I had one or two concerns about the proposed
review which I expressed to Patrick Bourke and Rodric Williams (for example,
that it could lead to a loss of privilege in documents considered by the review,
and that it could lead to difficulties for the POL legal team if a conflict arose
between Tim Parker and POL). WBD was not instructed in relation to the
commissioning of the Swift Review, its extent, nor how it was to be conducted.
Page 168 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
From my emails, I was asked on a couple of occasions to provide information
relating to the Mediation Scheme to feed into the review.
291. The outcome of the Swift Review was a lengthy report (POL00006355). Though
that report was dated 8 February 2016, I do not believe I received a copy of the
full report until 15 April 2016.% On receiving it, I do not recall any specific
thoughts that I had about it other than that its overview and assessment of the
Mediation Scheme broadly aligned with my own views of the position at that time,
and that I thought it was a measured document which reached sensible
conclusions. As I have touched upon above (at §216) and explain further below
(at §§299-302), there were one or two factual matters referenced in the report
which I had not become aware of whilst working on the Mediation Scheme,
including the fact that Deloitte had identified that certain “privileged users” at
Fujitsu may have the ability to edit or delete transaction data. I do not recall
having any conversations as to who the Swift Review should be provided to, and
I did not see a copy of Jonathan Swift QC’s instructions until 5 May 2016.24
292. Though I did not receive a full copy of the report until a couple of months after it
had been finalised, on 26 January 2016 POL contacted me to request my input
in implementing one of the (then draft) report's recommendations. This was that
POL should:
“cross-reference specific complaints about misleading advice from
NBSC [(Network Business Support Centre, the helpline for operational
issues arising in branch)] call-handlers with the possible employees who
provided that advice and consider their personnel files, where available,
232 See for example WBONO0000948.
233 WBON0000962.
234 POL00174470.
Page 169 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
for evidence as to the likelihood that the complaint may be well-
founded.’?°>
293. I was asked if my team of paralegals could analyse the 107 SPM cases that
made complaints about the Helpline during the Mediation Scheme with a view to
identifying those who had made particularised, rather than merely generic,
complaints. For context, it was not unusual for POL to make ad hoc or isolated
requests for assistance without explaining why or providing detailed background
(although in this instance, I was aware in general terms that the Swift Review
was underway and that a draft report had been delivered containing a number of
recommendations, of which this was one).
294. I responded that I could allocate three paralegals to the task, with the aim of
completing it that week.?°° I liaised with a trainee solicitor in my team to collate
the necessary information to do with the relevant cases.?°’ I reverted with the
results of that initial analysis on 1 February 2016.2°8 I expressed the view that 16
of the cases provided sufficiently detailed information that could be directly cross-
referenced with the NBSC call logs. 11 further cases alluded to a call happening
in a particular week or month (from which one could probably attempt to find the
details of the call from the call logs), though I took the view that this would require
a great deal of extra work (some of which would inevitably be guesswork) and
therefore I advised that we focus our efforts on the 16. I offered further paralegal
support to assist, in respect of these 16 cases, with (i) reviewing the NBSC call
logs to try to identify any relevant NBSC call reference ID numbers, and (ii) for
each identified call, reviewing the relevant POIR and CRR to determine what had
236 WBONO0000952.
236 WWBONO000954.
237 WBONO0000414.
238 WBONO0000955.
Page 170 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
been said about these calls and the issues raised. Patrick Bourke took me up on
that offer.?°9 Again, I asked a trainee to pull together those documents.”4°
295. I sent that second stage of the analysis to Patrick Bourke on 10 February 2016.741
The next step was to identify the call handlers by name, which I suspected would
have to be pulled from the call logging system in NBSC (and so, was an internal
job for POL). Mark Underwood responded on 17 February 2016 advising that I
should speak with Kendra Williams at NBSC to obtain the requisite information,
and raising a few points in relation to other cases within the NBSC complaints
analysis produced by WBD.24? I asked Paul Loraine (a solicitor in my team) to
help with this.24° He liaised with Kendra Williams to secure the case handler
information for every instance where we had a call reference number.2“4 My firm's
email records disclose that this involved a considerable amount of back and forth
between Paul and Kendra. Paul produced a final report on 5 May 2016.24
Looking at that report's conclusions, it appears that there were either no
performance complaints and/or concerns raised in relation to the call handlers in
the cases under consideration, or (in a small number of cases where a complaint
was made at the time) the data was insufficient for us to identify who the call
handler was. This closed off the recommendation, which was _ ultimately
Recommendation (7) in Jonathan Swift QC’s finalised report.
296. In addition, I can see from my email records that in February 2016, my firm
assisted POL in instructing Brian Altman QC to advise on the Swift Review's
238 WWBONO0000957.
240 WBONO0000415.
241 WBONO000958.
242 POL00239502.
243 WWBONO0000417.
244 WBONO0000419.
245 WBON0000990; POL00241260.
Page 171 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
recommendations which related to POL’s historic prosecutions (in particular,
Recommendations (1) and (2), which involved a review of the safety of false
accounting convictions that were procured by the dropping of theft charges
backed by insufficient evidence).*4° The nature of that assistance was limited;
Brian Altman QC was instructed by POL’s in-house legal team to carry out the
recommended review, and we simply prepared the bundle of documents to
accompany the instructions which POL had prepared. Paul Loraine helped in
completing that task.24”
297. I do not believe that I had any further involvement in the implementation of the
Swift Review recommendations until the group litigation. Separately, Deloitte was
engaged by POL in respect of the review’s IT-related recommendations, but I do
not recall being involved in this at this initial stage. I discuss Deloitte’s role during
the group litigation further below (in particular, at §§466-472, §§483-490, and
§§510-520). For completeness, I do not recall that I had any particular
involvement in the implementation of Recommendations (1) and (2), but Brian
Altman QC’s review was quickly subsumed within the group litigation and
adopted as a workstream therein (and again, I discuss this further below at
§§458-465).
(ii) The start of the group litigation
298. In order to contextualise the position going into the group litigation (which I deal
with in the sections that follow), I briefly address two matters relating to my receipt
of the Swift Review on 15 April 2016.
246 WWBONO000420.
247 WBONO0000420.
Page 172 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
299. The first was that I became aware of certain findings that had been made by
Deloitte about Fujitsu's remote access capabilities, which led Jonathan Swift QC
to recommend that further investigative work be undertaken by POL. The
background to this was that shortly before I was sent the Swift Review, I received
a copy of a summary report that Deloitte had produced for the POL Board in May
2014 (POL00028069) as part of its work on Project Zebra (the “Board
Summary”). I had not received the Board Summary previously despite being
provided with the Desktop Report on which it was based in August 2014, and
indeed I was not aware of its existence until early 2016.248 I became aware of it
in the context of helping POL with certain administrative aspects of its response
to a “section 17 request” by the CCRC,”49 namely, readying the documents
sought by the CCRC for submission and preparing a cover letter to go with
them.25° One of the documents mentioned in the request was the Board
Summary, so I sought it from Rodric Williams and he provided it on 8 March
2016.25
300. I cannot recall whether I read the Board Summary at that time or later when I
received the Swift Review. In any event, the combined effect of these two
documents was to make me aware of the concept of privileged users at Fujitsu
who had the ability to delete transaction data in the Core Audit Log, and
248 See above, §212.
249 Ie. a request under section 17 of the Criminal Appeal Act 1995.
280 Cartwright King was principally responsible for considering section 17 requests made by the
CCRC to POL and identifying the relevant documents or information. However, the actual
provision of documents to the CCRC was done by WBD through a data room hosted for that
purpose. The reason for this arrangement was that WBD had experience of operating data
rooms through an external eDiscovery provider (now known as Consilio), and so was better
placed than Cartwright King to establish the structure of the data room and operate it on a day-
to-day basis. As part of this work we kept a tracker of section 17 requests sent by the CCRC.
and POL’s progress in complying with them.
251 See my email of 10 February 2016 which stated, “/ don’t have a copy of the Deloitte board
report - do you have it?’ WBON0000960.
Page 173 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
potentially had the ability to do so in an undetectable manner. The Board
Summary highlighted in the executive summary (p.3) that “ijt is possible for
Fujitsu staff with suitably authorised privileged access to delete data from the
Audit Store”. It went on to explain (at p.6) that:
“Matter 3: Baskets of transactions recorded to the Audit Store are
complete and ‘digitally sealed’, to protect their integrity and make
it evident if they have been tampered with
[ud
Key Horizon Features ... are:
e Transactional data received into the central database is copied to
the Audit Store during an overnight process ...
e As part of this copying process, a ‘digital seal’ is applied to groups
of baskets ... The digital seal ... does not use cryptographic keys,
relying instead on the physical hardware control described below
to maintain the integrity of the digital seal itself.
e The Audit Store physically runs on ... specialist IT hardware which
protects data once it is written, preventing alteration of data in the
Audit Store. The digital seal codes are also written to the Audit
Store, thus providing a source for integrity checking that they
cannot be altered. If any data components within the relevant
group of baskets were to be altered, go missing or get added to,
then the digital seal for that group would be ‘breached’ and thus
the tampering could be detected. The configuration of the
physical hardware does however permit administrators to delete
data from the Audit Store during the seven year period, which was
a matter found to be possible and contrary to POL’s
understanding of this physical protection Feature. This could
allow suitably authorised privileged staff in Fujitsu to delete a
sealed set of baskets and replace them with properly sealed
baskets, although they would have to fake the digital signatures
[in an earlier section of the Board Summary, it was noted that no
documented controls had been identified which were designed to
“[p]revent a person with authorised privileged access to the digital
signing process from sending a ‘fake’ basket into [the] digital
signing process”.
Page 174 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
e Database access privileges that would enable a person to delete
Audit Store data are restricted to authorised administrators at
Fujitsu.
e Database access privileges that would enable a person to create
new entries, re-sealing it with a valid, (publically available) ‘hash’
are restricted to authorised administrators at Fujitsu.
We have not identified any documented controls designed to:
LJ]
e Prevent a person with authorised privileged access from deleting
a digitally sealed group of data and replacing it with a ‘fake’ group
within the Audit Store (which could still have a valid digital
signature, if they have access to keys, and a valid digital seal
created using a publicly available formula).’?52
301. The Swift Review commented that the Board Summary described this form of
remote access “more clearly ... than in any other document we have seen on
this subject’ (paragraph 139). It was reported that Deloitte “described this
functionality as resulting, in essence, from the level of security contained in
Horizon being a level down from the maximum” (paragraph 140), and that Fujitsu
“appearled] to accept that Deloitte’s interpretation is technically correct, but
emphasise[d] the wide range of security measures in the software, hardware and
environment which reduce the risk of interference” (paragraph 141).
302. As a result of reading the Board Summary and Swift Review, I came to
understand that certain authorised personnel at Fujitsu had the ability to delete
existing transaction data and replace it with new data, in effect changing the
transaction data recorded by branches. This was in direct contradiction to what
282 As explained above at §216, in the Project Zebra Desktop Report these matters were
referenced in Appendix 2, but I did not pick up on these references or appreciate their
significance at the time I read it in 2014.
Page 175 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Fujitsu had told me at various points through 2014 and 2015 as explained
above.253
303. The Swift Review concluded that in the light of the “consistent impression given”
by POL’s public statements that there was no ability to edit or delete Horizon
transaction data, “it [was] now incumbent on POL to commission work to confirm
... insofar as possible” how this capability was controlled and whether it had in
fact been used (paragraph 146). It was acknowledged that Fujitsu “properly,
stress that there is no evidence that any such action has occurred and that
likelihood of all the security measures being overcome is so small that it does not
represent a credible line of further enquiry’ (paragraph 141), and that the
recommended investigation into privileged user access was “most likely to be
wild goose chase” (paragraph 146). The recommendation to carry out an
investigation into privileged user access was Recommendation (5) of the
finalised report:
“(5) POL instruct a suitably qualified party to carry out a full review of
the controls over and use of the capability of authorised Fujitsu
personnel to create, amend or delete baskets within the sealed
audit store throughout the lifetime of the Horizon system, insofar as
possible.”
304. There were a number of other !T-related recommendations which the report
made and which later became relevant to my own work. These were:
“(3) POL consider instructing a suitably qualified party to carry out an
analysis of the relevant transaction logs for branches within the
Scheme to confirm, insofar as possible, whether any bugs in the
Horizon system are revealed by the dataset which caused
discrepancies in the accounting position of any of those branches.”
253 See paragraphs §§202 ff.
Page 176 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
And:
“(4) POL instruct a suitably qualified party to carry out a full review of
the use of Balancing Transactions throughout the lifetime of the
Horizon system, insofar as possible, to independently confirm from
Horizon system records the number and circumstances of their
use.”
And:
“(8) POL commission forensic accountants to review the unmatched
balances on POL’s general suspense account to explain the
relationship (or lack thereof) with branch discrepancies and the
extent to which those balances can be attributed to and repaid to
specific branches.
305. As noted above, at the time I received the Swift Review in April 2016 I was made
aware that POL was in the processing of commissioning Deloitte to carry out
Jonathan Swift QC’s IT-related recommendations. Specifically, Tim Parker was
doing this as part of his wider “Chairman's Review” into POL’s handling of SPMs’
complaints. Indeed, the reason why I was sent the Swift Review at this time was
so that I could advise on whether POL was likely to be able to assert legal
professional privilege over Deloitte's work; this is the second matter to which I
referred at §298 above.
306. This is reflected in an email which Rodric Williams sent to me and Gavin
Matthews on 15 April 2015, which refers to a telephone conversation he and I
had had (although I do not now remember that call), and continues:
“[bJefore Deloitte takes any further steps on its current engagement
could you please consider and advise as to whether anything
further can be done to strengthen Post-Office’s claim to privilege
over the work product which Deloitte will shortly be producing, e.g.
Page 177 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
by Bond Dickinson formally instructing the preparation of the work
product as POL’s external solicitors”?
307. By this point, we were aware of the threatened group litigation (see below, §309),
and the matters to be considered by Deloitte in response to the Swift Review
overlapped with the issues we anticipated were likely to arise in the group
litigation. I responded by email dated 19 April 2016 recommending that to
maximise POL's prospects of asserting privilege, Deloitte should be instructed
solely on the grounds of the upcoming litigation.2°°
308. WBD were therefore instructed to write to Deloitte explaining that the group
litigation had begun, that they were engaged by POL to provide expert advice in
connection with the litigation, and that WBD may therefore now provide
instructions to them.2°° We wrote this letter on 26 April 2016, which represented
the point where WBD began to take over the management of Deloitte’s work. I
address my involvement in the Deloitte’s work thereafter further below.
K. THE GROUP LITIGATION - GENERAL
(i) Introduction
309. I became aware in late 2015 that a claim against POL was likely to be
forthcoming. Proceedings were issued by 91 Claimants on 11 April 2016,
although the Claim Form was not served until August 2016 whilst the pre-action
254 WBON0000965. To this end, I was sent a copy of Deloitte’s "Change Note" to its previous letter
of engagement at the same time as Jonathan Swift QC’s report: WBON0000962;
POL00240675. The Change Note was the document that Deloitte used to set out the scope of
the proposed new work.
265 WBON0000339.
256 WBON0000984; WBON0000985; POL00242882.
Page 178 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
correspondence ran its course. WBD was engaged to represent POL in the
litigation, and I was the Partner in the firm with principal responsibility for the
conduct of POL’s defence (as mentioned towards the beginning of this statement,
I became a Partner in May 2016).
310. The group litigation was an enormous endeavour, occupying the majority of my
time from Spring 2016 until the beginning of 2020, and spanning multiple
hearings including two lengthy trials. At its peak, I was managing a team of 15
lawyers plus a substantial team of paralegals that varied in size as the volume of
work fluctuated.
311. In answering the Inquiry’s questions about these matters below, I have
endeavoured to maintain a chronological order so far as possible. This has not
been possible in every respect given that some of the topics identified in this part
of the Request are cross-cutting or relate to the same stages of the litigation as
others. Bearing this in mind, and in view of the ground that I have already
covered, I set out below a brief reminder of the structure I adopt in the sections
that follow:
311.1.This section (Section K) gives an overview of the group litigation and
answers the Inquiry’s Q58 to Q63, Q67 and Q70, broadly concerning my
relationships with key actors involved in the litigation, POL’s management
of the litigation, and WBD’s advice on strategy and tactics.
311.2.Section L (§§394-451) deals with the Inquiry’s Q64 to Q65, Q68 to Q69,
and Q71 to Q72, which broadly concern my/WBD’s early work (in
particular, during the course of 2016) in relation to preservation of
documents, early disclosure, and other forms of information-sharing.
Page 179 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
311.3.Section M (§§452-520) addresses the Inquiry's Q73 to Q75 and Q89,
covering the investigative and preparatory work undertaken to enable POL
to prepare its Letter of Response to the Claimants’ Letter of Claim (in July
2016) and its Generic Defence (in July 2017), as well as the report Deloitte
produced shortly after the Generic Defence was served, in September
2017.
311.4.Section N (§§521-694) covers the topic of disclosure (save for early
disclosure which is dealt with in Section L). In particular, in particular it
deals with the disclosure orders made at the CMCs from October 2017 to
June 2018 and the approach to disclosure thereafter, answering the
Inquiry’s Q58.4, Q76 to Q88, Q90.1, Q91, Q95.1 and Q99.
311.5.Section O (§§695-768) answers the Inquiry’s Q90 and Q92 to Q94 of the
Request, concerning POL’s preparation for the Common Issues Trial
which took place over 15 days in November and December 2018 (Q90.1
and Q91 being dealt with in the preceding section on disclosure).
311.6.Section P (§§769-912) answers the Inquiry’s Q95 to Q102 of the Request,
concerning POL’s preparation for the Horizon Issues Trial which took place
over 21 days between 11 March and 2 July 2019 (albeit that Q95.1 and
Q99 are dealt with in Section N on disclosure).
311.7.Section Q (§§913-989) answers the Inquiry’s Q103 to Q118, concerning
POL's response to the Common Issues Judgment including, in particular,
the Recusal Application.
Page 180 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
311.8.Section R (§§990-1006) summarises events after the conclusion of the
Horizon Issues Trial, and therefore briefly deals with the Inquiry’s Q119 to
Q120.
(ii) Overview of the group litigation
312. It would not be practical for this statement to cover the whole history of the group
litigation. I am also conscious that the Inquiry will be familiar with much of the
history of the litigation. However, I do believe that it is helpful to consider the
Inquiry's questions and my responses with the following overview of the key
events in mind.
313. Having issued the Claim Form on 11 April 2016, the Claimants set out a summary
of their allegations in a Letter of Claim (“LOC”) sent on 28 April 2016. POL’s Letter
of Response (“LOR”) was sent on 28 July 2016. The Claimants’ LOC invited POL
to accede to the making of a GLO, and the parties corresponded about this whilst
the LOR was being prepared (which included me discussing the matter directly
with James Hartley, the Partner at Freeths who had conduct of the Claimants’
case).”5” An unusual feature of the group litigation was that each Claimant's
circumstances and the nature of their complaints against POL were in many
respects quite different, such that it was not always easy to identify (at least with
precision) the common issues between them. Despite this, POL agreed in
principle to the making of a GLO and this was communicated to the Claimants,
including in POL’s LOR dated 28 July 2016.
287 WBON0000336.
Page 181 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
314. Just before the LOR was sent, on 26 July 2016, 107 Claimants were added to
the Claim Form by way of pre-service amendment, and on 29 July 2016 the
Claimants formally applied for a GLO. Following that letter the parties continued
to correspond with a view to agreeing the detail of the proposed order. There was
a fairly lengthy wait for a GLO hearing to be listed and that was ultimately
scheduled for 26 January 2017 (the “GLO Hearing”), with the Claimants
providing draft Generic Particulars of Claim (“GPOC”) in December 2016.
315. The GLO was made at the GLO Hearing before Senior Master Fontaine, and the
approved order sent to the parties on 21 March 2017.58 At that stage, a
Managing Judge for the litigation was yet to be appointed but a first CMC was
listed for the first available date after 18 October 2017 (and ultimately took place
on 19 October).?59
316. Between the GLO Hearing and the first CMC, the Claimants served their Generic
Particulars of Claim on 23 March 2017 and their Amended Generic Particulars of
Claim on 6 July 2017. POL then filed its Generic Defence and Counterclaim on
18 July 2017. Producing POL's defence required an intensive period of work but
there were few other directions for POL to comply with at this time. The original
cut-off date for new Claimants to join the litigation was set as 26 July 2017 by
Senior Master Fontaine, and by a second Claim Form issued on 24 July 2017
324 additional Claimants joined the action. The Claimants were also required by
Senior Master Fontaine’s directions to produce Schedules of Information
288 WBON0001674.
289 This hearing was briefly resumed on 25 October 2019 to address Counsel's availability for the
Common Issues Trial, which led to Mr Justice Fraser's judgment [2017] EWHC 2844 (QB),
addressed further below at §§385 ff.
Page 182 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(“SOls”) (which were to take the place of individual pleadings in their case), and
they began serving these in tranches from June 2017 until December 2017.
317. The shape and pace of the litigation then changed significantly at the first CMC
on 19 October 2017. Mr Justice Fraser decided that a first trial would be listed
for a 20-day period commencing on 5 November 2018, the purpose of which was
to determine 23 Common Issues relating to the legal relationship between POL
and its SPMs, both where this was governed by the SPMC and where the later
Network Transformation Contracts (“NTC”) applied. This trial became known as
the Common Issues Trial or “CIT”. Mr Justice Fraser also decided that a further
trial would be held in March 2019, the issues for which were not set at this time.
318. Other key directions given by Mr Justice Fraser at this stage included:
318.1.The cut-off date for the GLO was extended to 24 November 2017. Taking
into account additional Claimants who joined the litigation by this point and
discontinuances by some Claimants, the total number was around 550.
318.2.The parties had to exchange Electronic Disclosure Questionnaires
(‘EDQs”) by 6 December 2017.
318.3.Lead Claimants for the CIT had to be agreed by 23 February 2018 with
subsequent directions for them to plead their cases in more detail.
318.4.POL was to provide some initial tranches of disclosure in relation to
prospective Lead Claimants for the CIT in January 2018.
318.5.The parties were to prepare a statement of agreed facts in relation to the
Common Issues by 29 June 2018.
318.6.Witness statements for the CIT were to be filed by 11 August 2018.
Page 183 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
318.7.The Court also gave permission, at this stage, for each party to rely on an
IT expert “in relation to the operation and accuracy of the Horizon system”;
The Claimants’ IT expert was Jason Coyne and POL’s was Dr Robert
Worden. POL agreed to use reasonable endeavours to facilitate a visit by
Mr Coyne to Fujitsu’s Bracknell Office to inspect the KEL.76
319. This began an intense period of activity because we had just one year to prepare
for a significant trial, with another to follow quickly thereafter. We had to expand
the size of the WBD team to meet the Court's deadlines and the work became
more intensive and time pressured. Tony Robinson QC (who had been retained
shortly after receipt of the LOC as POL’s Leading Counsel) was not available for
the CIT, so David Cavender QC was engaged to lead on the Common Issues
with Tony Robinson QC being held to deal with the second trial.
320. Further directions in relation to CIT disclosure were made at a CMC on 2
February 2018.
321. Ata CMC on 22 February 2018:
321.1.Further directions relating to generic disclosure for the CIT were made,
requiring substantial disclosure to take place by 18 May 2018.
321.2.Mr Justice Fraser decided that the second trial in March 2019 would deal
with issues relating to the Horizon system (and consequently this trial
became known as the Horizon Issues Trial or “HIT”). The effect of this was
that the GLO issues were determined in stages by a series of thematic
260 WBONO001685.
Page 184 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
trials as opposed to, for example, a lead or test claimant approach (being
the approach that POL had proposed).
321.3.Disclosure orders were made in relation to the HIT, to be complied with by
13 April and 18 May 2018 respectively.
321.4.Mr Justice Fraser gave directions in relation to expert evidence for the HIT,
including for the Claimants to provide an outline of their allegations in
relation to the Horizon system; for Mr Coyne to provide his first expert
report (“Coyne 1”) by 14 September 2018 (subsequently varied to 12
October 2018); for Dr Worden’s first report (“Worden 1”) to be served by
2 November (subsequently varied to 7 December 2018); and for the
experts to exchange supplemental reports (“Coyne 2” and “Worden 2”,
respectively) in January 2019.26"
322. A further CMC took place on 5 June 2018. At this hearing:
322.1.POL expressed concern about the scope of the factual matrix that would
be in issue at the CIT, in light of indications from the Claimants that they
intended to rely on a wide range of evidence going to contractual
performance and breach of duty by POL, in support of their case on the
questions of construction, incorporation and implication of contractual
terms that made up the Common Issues. Mr Justice Fraser did not make
any directions in relation to this but gave the Claimants an oral warning
about adducing inadmissible evidence and noted that POL might apply to
strike out such evidence if it was served.
261 POLO0117925.
Page 185 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
322.2.Mr Justice Fraser made further orders for disclosure in relation to the HIT,
to be complied with by 17 July 2018.
322.3.The parties were ordered to file factual witness statements for the HIT by
14 September 2018 (subsequently amended to 28 September 2018).
322.4.POL was ordered to file supplementary witness statements by 16 October
2018 (subsequently amended to 16 November 2018), and the Claimants
by 14 December 2018 (subsequently amended to 17 January 2019),
respectively.262
323. In the event, when the Claimants’ factual evidence for the CIT was served in
August 2018, we took the view that it contained large amounts of material which
was inadmissible, being irrelevant to the issues which the Court had to decide at
the CIT (though it would have been relevant and admissible at future trials on
breach and causation). POL therefore applied on 5 September 2018 to strike out
parts of this evidence, which application was refused on 17 October 2018.
324. The CIT took place across 15 non-consecutive days in November and December
2018, with the resulting Common Issues Judgment being circulated in draft on 8
March 2019 (one working day before the start of the HIT) and formally handed
down on 15 March 2019.
325. The HIT then took place across 21 non-consecutive days between 11 March and
2 July 2019, with an adjournment during that time whilst the Recusal Application
issued by POL on 21 March 2019 was dealt with. Mr Justice Fraser’s judgment
on the Horizon Issues (the “Horizon Issues Judgment’) was handed down on
262 POL00120352.
Page 186 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
16 December 2019, by which time the parties had just settled the litigation
following a nine-day mediation.
326. For completeness, on 31 January 2019 there was a further CMC at which Mr
Justice Fraser fixed a Further Issues Trial to commence on 4 November 2019
(although that was subsequently postponed and never held because the litigation
was settled).26° Although it never reached this point, I anticipated that there would
likely have been at least a fourth, and maybe a fifth, trial before the Court had
addressed all the key issues in dispute. There would then likely have been some
exercise of applying those findings to the c.550 individual cases because there
was no one issue in the group litigation that was dispositive of all the claims. At
the time group litigation settled, it was in my opinion less than halfway through
its total course.
327. The impact of the above was that, from early 2018, we were preparing for two
significant trials simultaneously, which concerned wide-ranging issues, the latter
of which (the Horizon Issues) was still taking shape. We relied on a large team
of Counsel and expanded the WBD team to ensure we had adequate resource.
However, there were occasions where my time had to be focussed on one or
other of the trials. Particular pinch points included:
327.1.The deadline for the initial tranche of factual witness statements for the
HIT (28 September 2018) fell during a period of intense preparations for
the CIT which was due to commence on 5 November 2018.
327.2.The deadline for POL's supplementary witness statements for the HIT fell
in the middle of the CIT.
263 WBONO001669.
Page 187 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
327.3.Work on Worden 1 was ongoing during the CIT, being served on 7
December 2018.
327.4.The Common Issues Judgment was circulated in draft immediately before
(and was formally handed down during) the HIT, meaning that we needed
to make decisions on POL's response to that judgment whilst I was in trial.
328. A further impact of the approach taken was that it allowed less time for reflection
and settlement discussions than might otherwise have been the case. For
example, POL was keen to hold a mediation in February 2019 (having expected
the Common Issues Judgment by then) but the Claimants ultimately wanted to
wait until after the HIT. It was difficult to find a window for a mediation and it is
possible that a more conventional lead or test claimant approach, entailing all of
the issues in relation to those cases being ventilated, might have enabled
settlement discussions to take place at an earlier stage.
(iii) Instructions and relationships (Q58 to Q59)
329. In answer to Q58.1 of the Request, I received instructions on a day-to-day basis
from Rodric Williams and Jane MacLeod (until Jane McLeod was effectively
replaced by Herbert Smith Freehills, “HSF”, in April 2019). The Postmaster
Litigation Steering Group (the “Steering Group”) was set up by POL early on in
the process to make key strategic decisions in the litigation and to sign off
important correspondence and documents. I describe the Steering Group further
below (§§336 ff), but in short it included representatives from all relevant areas
of POL (legal, branch network, IT, finance, communications, etc). For most of the
group litigation, significant pieces of advice which WBD gave were mainly
submitted to the Steering Group in the form of Steering Group papers. The
Page 188 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Steering Group generally met monthly and sometimes it met more often than
this. I did not have any reporting line beyond the Steering Group. Any such
reporting line from the Steering Group to senior management was managed by
Jane MacLeod and the Chair of the Steering Group (Tom Moran, and later Jane
MacLeod, when Tom Moran left).
330. In around early 2018, POL set up a Board Subcommittee of the POL Board to
oversee the litigation (the “Board Subcommittee”). This was when I started to
have direct contact with POL Board members as I was invited to attend some of
their meetings (save for a few other occasions set out below at §345 and §347).
The Steering Group was still operational at this time, but more important
decisions were put to the Board Subcommittee.
331. By Q59 I am asked to comment on my relationship with POL’s representatives
during the litigation. My experience was that POL’s senior management was
actively involved in overseeing the conduct of the litigation through the Steering
Group and later the Board Subcommittee, and that they asked appropriately
challenging questions. Where senior management disagreed with a particular
piece of advice, they said so (see, for example, the minutes of the Steering Group
meeting on 6 December 2017 which record members of the Steering
Committee’s views on WBD’s advice on settlement).2°* When the Board
Subcommittee was established, I found that the Board Subcommittee was
similarly engaged and actively involved in setting the strategy.
332. In my view, POL had a reasonable sized in-house legal team for the purposes of
the group litigation, because it chose to outsource the day to day running of the
264 POL00251998.
Page 189 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
litigation. My impression was that Rodric Williams worked on this case nearly full
time, although I was aware that his capacity was stretched. About halfway
through the group litigation, WBD hired and placed on fulltime secondment to
POL a senior in-house litigation lawyer (Ben Beabey); my understanding was that
Ben covered all non-group litigation work so that Rodric Williams could focus his
efforts on the group litigation. Rodric Williams was an experienced civil litigation
lawyer with, I believe, limited criminal law experience.
333. As mentioned above, we instructed Tony Robinson QC as POL's Leading
Counsel for the group litigation in late May 2016 and I worked closely with him
throughout the litigation. Later we also instructed David Cavender QC to
represent POL in the CIT because Tony Robinson was unavailable. My
relationship with POL’s Counsel was the same kind of relationship that I have
with all Counsel that I work with, viz. regular interaction and discussion about
most issues, where Counsel and I both suggest ideas and challenge each other's
thinking. We worked well together as a team and had a good rapport. A
substantial part of my role was ensuring that all the various workstreams and
action points were running on time and coming together. I would leave Counsel
to address points of detail that were properly their remit (for example, pleadings,
submissions at hearings, preparation for cross-examination at trial, etc.)
334. In terms of my reliance on the advice of Counsel as to (a) general litigation
strategy (b) POL’s approach to disclosure (c) the preparation of witness evidence
and (d) the recusal application:
334.1.As to (a), I discussed the general litigation strategy with Counsel
frequently. Counsel fed in on all aspects of the litigation strategy at each
stage, and their advice was reflected in papers submitted to the Steering
Page 190 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Group and was regularly fed to POL orally and by email. Counsel regularly
met with members of the POL in-house legal team, and also provided
formal advice directly to senior management on a number of occasions.
On 29 September 2017 (shortly before the first CMC), Tony Robinson QC
met with Paula Vennells and Alisdair Cameron at POL to discuss the
overall strategy (see further below, §349). When David Cavender QC was
first instructed, WBD asked him to review the entire litigation strategy and
critique our approach to the case (see §391 below).7°° On 15 May 2018,
the full Counsel team at that stage (David Cavender QC, Tony Robinson
QC, Owain Draper and Gideon Cohen) provided an advice on the
prospects of success at the CIT.?® In May 2018, at a meeting of the Board
Subcommittee, David Cavender QC and Tony Robinson QC gave a
detailed presentation on the overall merits of POL’s case, as well as advice
on the general strategy and prospects of settlement.7°”
334.2.As to (b), Counsel were involved in advising on POL’s approach to
disclosure both in terms of the scope of disclosure orders that were made
during the course of three CMCs in early 2018 (see §318 and §§320-321
above) and inputting on key disclosure points. However, advice on
preservation of documents and day-to-day advice on disclosure was given
by WBD without reference to Counsel.
334.3.As to (c), Counsel were involved in all material decisions relating to witness
evidence including decisions about who to call, who not to call and what
265 POL00251957. He also prepared a brief advice in January 2018, seeking to identify the points
which he considered were the Claimants’ strongest: POL00252996.
266 POL00270841.
267 WBONO001688.
Page 191 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
evidence should be given by a particular witness. To the best of my
recollection, I believe that the majority of draft witness statements, if not
all of them, were also reviewed by Counsel. At §§730-737, §747 and §749,
and §§789-790 of this statement, I address in more detail the Counsel
team’s role in relation to the aspects of the witness evidence about which
the Inquiry has asked more specific questions.
334.4.As to (d), I address the Recusal Application below in Section Q, but in
summary the advice on the merits of this application was primarily
provided by Counsel. In the first instance, David Cavender QC advised
that large parts of the Common Issues Judgment made findings which
were outside the ambit of the issues which fell to be determined. In so
doing, Counsel thought Mr Justice Fraser had made findings that
prejudged matters that were properly within the scope of later trials. This
was the essential genesis of the recusal application, and (as I explain
further below) POL subsequently instructed Lord Neuberger to advise on
the application and Lord Grabiner QC to advise on and presentit. I agreed
with the idea of making the Recusal Application, but I had no experience
of such applications so the strategy, and advice on its merits, was led by
the heavyweight Counsel team that had been brought in for that purpose.
335. In terms of my relationship with Freeths and the Claimants’ Counsel, I always felt
able to phone James Hartley, the Partner at Freeths. We also held a meeting
with Freeths to discuss the scope of disclosure on 22 December 2017 (see
below, §538) and a further meeting with Freeths and the IT experts on 11 April
2018 to help shape the approach to the Horizon Issues (§544). My conversations
with James Hartley were always cordial, thought we often had different views on
Page 192 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
various issues. Generally, I felt those conversations helped find some common
ground. However, I always had the feeling that Mr Hartley was reluctant to make
decisions without first getting sign off from his Counsel team which did inhibit the
progress of our discussions. I had no direct contact with the Claimants’ Counsel
save in the margins of hearings and trials.
(iv) POL’s management of the litigation (Q60 to Q63)
The Steering Group and Board Subcommittee
336. As noted above, the Steering Group was formed early on to oversee the litigation
and make strategic decisions on behalf of POL. My experience of POL is that it
often formed committees to oversee material activities in its business. I do not
recall specifically how the Steering Group came to be formed, but when it was, I
was not surprised. I recall discussing with Jane MacLeod and Rodric Williams at
a very early stage how important it was for representatives from across the
business to be directly involved in giving instructions, and for this not to become
a purely ‘legal’ issue whereby the in-house legal team was solely responsible for
directing the external lawyers. Jane MacLeod and Rodric Williams agreed; they
generally held the view that the in-house legal team’s function was to advise and
it was for the business to make decisions. I do not know to what extent this
conversation was fed back to POL’s senior management or whether this was the
genesis for the formation of the Steering Group.
337. Members of the Steering Group who regularly attended meetings were: Tom
Moran, Angela van den Bogerd, Patrick Bourke, Mark Davies (and after Mark left,
Mel Corfield), Rob Houghton or Catherine Hamilton (from the IT department),
Nick Beal, Tom Wechsler, a representative from the finance team which changed
Page 193 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
over time, Jane MacLeod, Rodric Williams, Mark Underwood and myself. The
Chair of the Steering Group was Tom Moran (until he left and Jane MacLeod took
over). Over time other attendees from around POL's business joined various
meetings.
338. The principal purpose of the Steering Group was to provide POL’s instructions
as to the conduct of the litigation. In practice, Mark Underwood was responsible
for organising the Steering Group.©° There were typically monthly meetings and
fortnightly calls, in advance of which an agenda would be circulated by Mark
Underwood.”®° Briefing papers would also be provided in various formats
(including ‘Decision Papers’, ‘Discussion Papers’, and ‘General Updates’),
depending on the particular subject which the Steering Group needed to be
briefed on. By way of further explanation, the purpose of Decision Papers was to
advise on different decisions which needed to be made by POL, to enable POL
to give instructions (for example see: “Should Post Office undertake further work
to preserve relevant documents”.2”° “Should Post Office change the way it deals
with Active Claimants?”;2"' “Should Post Office pay Fujitsu to employ additional
staff to extract transaction data’);2”* whereas Discussion Papers were to set out
points for discussion in relation to particular issues (for example, see: “Next 12
months’).2’5 The purpose of the ‘General Updates’ papers is self-explanatory.2”
Other papers provided included factual briefing notes on the relevant aspects of
the litigation process.2’”° After the initial few meetings, a list of actions was
268 WBON0000511.
269 POL00139298; POL00243195.
270 POL00139297.
271 POL00139479
272 POL00251593.
273 POLO00251596.
274 POL00261175.
275 POL00261176; POL00261172; POL00259673.
Page 194 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
captured and circulated to attendees which would then be considered at
subsequent meetings to monitor progress but this process stopped at some point
during the litigation — I do not recall why. From reviewing my firm's files, I can see
that WBD submitted over 100 Steering Group papers during the course of the
group litigation.
339. The relationships between Steering Group members were as one would expect,
professional and cordial. I was not aware of any tension between members of
the Steering Group. Some members were better prepared for meetings than
others. In terms of the nature and extent of the discussions that the Steering
Group had on issues such as general litigation strategy, disclosure, and the
preparation of lay and expert witness evidence, that varied depending upon the
issue. I would routinely provide a verbal synopsis of each paper provided to the
Steering Group. Sometimes Decision Papers were presented which contained
recommendations from WBD that could be agreed with little debate and others
gave rise to substantial debate. In general terms I found that the level of debate
and engagement was appropriate to the nature and complexity of the issues at
stake and as I have said, I found that the Steering Group generally asked
appropriate and probing questions.
340. WBD did not take minutes of the Steering Group meetings and I cannot say
whether POL did; I recall that there was an action list produced following the early
meetings. For the purposes of preparing this statement, searches have been
conducted for copies of such minutes and none have been identified. Typically, I
would relay by email or telephone the decisions of the Steering Group to the
relevant person in my team following each meeting so they could then take the
matter forward as instructed. The above paragraph (and my evidence elsewhere
Page 195 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
in this statement as to what views were expressed at Steering Group meetings)
therefore reflects the best of my recollection, assisted by reviewing the relevant
Steering Group papers where available, and my email records. By way of
illustration of how discussions among the Steering Group generally proceeded, I
have identified a Decision Paper setting out a proposal for the March 2019 trial
and a long-term strategy for the group litigation.2”° Given that the decision was
urgent, there was no meeting, and members provided their views in writing by
email. The comments provided by the various Steering Group members are
indicative of the approach that was taken at the meetings themselves, see for
example the comments from Tom Moran,”’ Mark Ellis,2”8 Patrick Bourke2’? and
Nick Beal.2°
341. Later, in or around March 2018, the Board Subcommittee was set up to take the
major strategic decisions in the litigation. I do not know why this was, but at this
stage WBD were moving into the phase of preparing for the CIT and HIT in
earnest. I attended a handful of these meetings. From this point, it felt to me as
if the Board Subcommittee was more directly making decisions on the overall
strategic direction of the litigation (with the Steering Group still taking the material
tactical decisions on a regular basis). However, this may have only been my
perspective as I only infrequently met with General Executive or Board members
(other than Jane MacLeod who sat on the General Executive). It may have been
that they were much more actively involved from the outset in directing the
litigation in ways that were not visible to me.
276 POL00024436; POL00252205.
277 POL00024281; POL00252201; WBONO0000188.
278 WBON0000328.
278 WBONO000171.
280 POL00024278.
Page 196 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Meetings with POL Board members
342. Generally, I had limited contact with members of the POL Board, which I set out
below based on a review of my emails and Outlook calendar.
343. Save for a meeting on 30 October 2018, I do not believe that I attended any full
Board meetings. My firm’s records indicate that I attended the 30 October
meeting for around half an hour with David Cavender QC, for the purpose of
providing the Board with a general update ahead of the CIT (which was due to
start the following week).26" I do not recall this meeting. I did on a few occasions
input into POL Board papers. On 18 September 2017, I gave views on a POL
Board paper about litigation options.”°? This paper was, in essence, a covering
note to the litigation options table WBD had already produced for the Steering
Group. On 25 July 2018, I provided some input into a Board paper on
contingency planning.
344. As stated above, I attended a handful of Board Subcommittee meetings. In April
2019, HSF were appointed and my attendance at Board Subcommittee meetings
reduced from that point. I was asked to provide comments on Board
Subcommittee papers dated 11 November 201978 and 9 December 2019.785
345. I occasionally (from memory, it may have only been one or two times) attended
General Executive meetings, of which Paula Vennells was a standing member
(as well as being a member of the Board). For example, from my emails I can
281 WBON0001341.
282 POL00024633 and POL00117761.
283 POL00024235 and POL00358137.
284 WBONO0001658 and POL00288584.
285 WBON0001663 and POL00289960.
Page 197 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
see that on 7 November 2017, I was scheduled to speak for 10 minutes at a
General Executive meeting.7°°
346. To the best of my recollection and based on document searches, I prepared a
paper on the Generic Defence for the General Executive on 7 July 2017,78” and
on 8 January 2020 I inputted into a possible General Executive paper on
operational issues in light of the Common Issues Judgment (that paper has not
been included with this statement as I am unsure whether it falls within POL's
privilege waiver).
347. I can also see from my calendar that I attended other meetings and calls with
POL Board members outside of formal General Executive and Board
Subcommittee meetings (some in person and others by conference call) on the
following occasions:
347.1.0n 7 March 2017 I had a meeting with Alisdair Cameron, Deloitte and
others titled ‘Allegations Made in the Group Litigation re POL's Operation
of Suspense Accounts’.
347.2.0n 17 July 2017, I had a meeting Alisdair Cameron, Deloitte and others
titled ‘POL's Defence: Suspense Account Wording’ .
347.3.0n 18 September 2017, I had a meeting with Paula Vennells, Jane
MacLeod, Alisdair Cameron, and others titled ‘GE Briefing PSLG CMC’.
347.4.0n 22 September 2017, I had a meeting with Paula Vennells, Alisdair
Cameron, Jane MacLeod and others, titled ‘PSLG Decision Meeting’.
286 WWBONO0000510.
287 POL00249671 and POL00249674.
Page 198 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
347.5.On 20 October 2017, I had a meeting with Paula Vennells, Jane MacLeod,
and others titled ‘Postmaster Litigation briefing call’.
347.6.On 25 April 2018, I had a meeting with Tom Cooper, Jane MacLeod, and
others titled ‘Sparrow’. I recall this was a background briefing on the
litigation for Tom Cooper who had recently joined the POL Board.
347.7.0n 5 October 2018, I had a meeting with Paula Vennells, Jane MacLeod,
Alisdair Cameron, and others titled ‘GLO Contingency Planning’.
347.8.On 29 October 2018, I had a meeting with Paula Vennells, David Cavender
QC and others titled ‘Call re POL Litigation case’. This was in preparation
for the Board meeting the following day to which I have referred above at
§343.288
347.9.On 11 April 2019, I had a meeting with Tom Cooper, Alisdair Cameron,
Jane MacLeod, Norton Rose and others, called ‘Postmaster Litigation
Briefing’.
347.10. On 13 May 2019, I had a meeting with Alisdair Cameron, Ben Foat, HSF
and others titled ‘Group Litigation — Implications of Recusal Judgment’.
347.11. On 17 July 2019, 24 July 2019 and 30 July 2019, I had meetings with
Ben Foat, Tom Cooper, HSF and others variously titled ‘Group Litigation
Call’ and ‘GLO — Led Claimant Case Summaries’. I recall that these were
calls to brief Tom Cooper on the facts of the lead cases selected in the
group litigation.
288 WBONO0001341.
Page 199 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
348. Other than set out above, I do not have any recollection of the specifics of any
these meetings. There is also one return for a search of my Outlook calendar for
Alisdair Cameron, which refers to a ‘Call — Al Cameron/Andy Parson’ on 17 April
2019. I recall that Alisdair Cameron phoned me to say that Jane MacLeod was
leaving and that HSF was being brought in.
349. Q63 refers me to an email which Amy Prime sent me on 28 September 2017
(POL00006384) and an email which I sent to Jane Macleod on the same date
(POL00006499). Both of these emails relate to a meeting that occurred between
Tony Robinson QC and Paula Vennells on 29 September 2017. I could not attend
this meeting due to a family matter.
(v) Advice on strategy and tactics (Q58.2, Q58.7, Q66 to Q67, Q70)
General advice on litigation strategy
350. The first strategic question in the litigation was whether POL would oppose the
making of a Group Litigation Order. Tony Robinson QC's advice at a conference
held on 9 June 2016 was that POL should agree to a GLO, but he had some
concerns about its scope and the terms that would need to be negotiated with
the Claimants or decided on by the Court (see below, §§426-427). The GLO
application was heard on 26 January 2017. After that, no further Court hearings
took place until October 2017, during which time the parties produced generic
statements of case, the group litigation was open to new Claimants to join, and
each Claimant also had to serve an SOI containing partial details of their claim.
More than 320 Claimants joined the action in July 2017 and by the time the GLO
closed in December 2017 there were around 550 Claimants, with the SOls being
served in tranches between 20 June and 15 December 2017. It was therefore
Page 200 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
not until mid to late 2017 that number of Claimants and the shape of their claims
came into focus and a strategy for the rest of the litigation could be properly
considered.
351. The first CMC was listed for 19 October 2017, and this prompted focus on the
future course of the litigation. With reference to Q58.2 and Q66, in terms of
general litigation strategy, Tony Robinson QC and I considered that there were
several broad strategies which POL could potentially pursue in relation to the
group litigation, which we had previously discussed on numerous occasions over
the course of the preceding months. These were outlined in a paper for a
Steering Group meeting dated 11 September 2017 (POL00006380), and were
then drawn into an Options Paper which was prepared for a meeting of the
General Executive on 18 September 2017 (see above, §347.3). Those options,
and the recommendations in relation to each of them as of September 2017,
were presented in the Option Paper as follows:89
“1, Focus on Horizon
Push the Court to address at an early stage whether Horizon is robust
and accurately records branch transactions.
Recommendation: We do not believe it is possible to address this issue
without first establishing Post Office's legal obligations in relation to
Horizon (see Option 2)
2. Focus on contractual issues
Push the Court to address at an early stage whether the postmaster
contract is fair and whether it supports Post Office's current operating
practices.
Recommendation: This is our recommended approach in conjunction
with Option 3.
289 POL00250513.
Page 201 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
3. Focus on weak claims
Ask the Court to strike out Claimants who are facing legal and procedural
problems, such as their claims being out of time, having previously
signed settlement agreements or generally having very weak claims on
their own facts.
Recommendation: We do not believe that a Court would focus on these
Satellite issues in insolation as this would not tackle the major issues at
the heart of litigation. They could however be addressed in conjunction
with Option 2.
4. Settle now
Try to agree a settlement now that closes down the litigation at an early
stage.
Recommendation: This option is not recommended as it would result in
Post Office having to pay significantly over the odds.
5. Attrition
Stretch out the litigation process so to increase costs in the hope that the
Claimants, and more particularly their litigation funder, decide that it is
too costly to pursue the litigation and give up.
Recommendation: This option is not recommended as we believe the
pressure on, and cost to, Post Office would become unbearable before
the Claimants gave up”.
352. Consistently with my duty as a litigator both to the Court and to my client to act
in their best interests, we presented a full range of options covering a wide
spectrum of possible strategies. Our recommended approach, as the above
excerpt makes clear, was to pursue Options 2 and 3 in tandem.
353. As for the recommendations on Options 1 and 2 (suggesting an emphasis on
resolving contractual as opposed to Horizon issues), the essence of POL's legal
case was that if an SPM submitted their accounts without contemporaneously
raising any concern about them, they were bound by those accounts and liable
to pay to POL the sums of cash shown in them. I considered the contractual
Page 202 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
issues to be a priority because the Court's findings on the relationship between
SPMs and POL would underpin who bore the burden of proving the root cause
of a shortfall in a branch. The contractual issues would also set the basis for
assessing other points, for example, they would determine the extent and content
of POL’s contractual obligation to provide training and support as well as POL’s
obligations in terms of providing Horizon and the wider branch accounting
processes. The issues around Horizon could not, in my opinion, be meaningfully
framed or determined until the contractual position was established. This is why
WBD and Counsel advised POL to push for a trial on the contractual issues to
be held before moving onto other matters in dispute.
354.1 recall that the Steering Group was not initially predisposed to putting the
contractual issues at the heart of the case because they felt that POL had a
stronger case on Horizon and they did not want to be seen to be running away
from that issue. However, the legal team explained and advised that the logical
way to proceed was to address the contractual issues first because this was the
foundation for the other issues, including in relation to Horizon.
355. As for the Option 3 recommendation, to seek to strike out weak claims (in
conjunction with Option 2), there were various types of Claimants whose claims
were liable to be struck out because (i) they were dissolved companies, (ii) they
were companies incorporated after they were said to have been engaged as
SPMs, (iii) they were bankrupt, and/or deceased, and/or had entered into
settlement agreements with POL, or (vi) the claims were time-barred. As the
paper cited above explained, it was “quite proper’ that the potential weaknesses
and/or deficiencies with these claims be addressed early on, “as would be the
case in any other piece of litigation’. The paper also highlighted that the risk
Page 203 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
associated with this approach was that it would require satellite issues to be run
in parallel to the main claim, which could be seen by the Court to be too
burdensome for it to manage. Further, it would have limited effect on the overall
dynamic of the case.
356. Regarding Option 4, I explained that I expected a settlement without having
progressed any matter to trial would likely cost in excess of £40m. This was
based on my expectation that the Claimants litigation funder would want a
significant return on its investment before any compensation would reach the
SPMs. I give an overview of the approach taken to settlement throughout the
litigation below, at §§361-372.
357. As for Option 5, as the paper makes clear, this was nota strategy I recommended
because the cost to POL would be too high before the Claimants would feel the
impact of this approach. Insofar as it is to be suggested that this approach — or
indeed any approach which had the effect of applying pressure to the Claimants
— would have been inappropriate (or that any individual steps POL took that put
pressure on the Claimants were inappropriate), I highlight that applying a
reasonable degree of pressure to one’s opponent, and/or pursuing strategies
which have the effect of applying pressure to one’s opponent, are part and parcel
of an adversarial system of litigation. It was proper to include this so POL could
see the full range of options. However, at no stage did I recommend that POL
adopt this strategy and at no stage did POL instruct me to adopt such a strategy.
358. More generally, as explained above, where it was consistent with my duties to
the Court, my client, and my professional obligations for WBD to advise on
290 POL00250466.
Page 204 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
approaches which had the effect of applying pressure on Freeths and the
Claimants' litigation funder, these approaches would be explained to POL as
advantages of a particular step or action. However, at no stage did I advocate
taking a step purely for this effect. Where this factor infrequently arose, there was
always an overarching meritorious reason for recommending a particular course
of action, a by-product of which may have been to place pressure on the
opposing legal team. This is expected in adversarial litigation of any nature and
especially litigation of this scale.
359. The sentence, “Our target audience is therefore Freeths, the funder and the
insurers who will adopt a cold, logical assessment of whether they will get a pay-
out, rather than the Claimants who may wish to fight on principle regardless of
merit’ did not encapsulate WBD’s advice on how POL should approach the group
litigation (cf. Q66.3). WBD’s advice was to focus on the contractual issues, win
on those issues, and then to settle the balance of the case (the broad approach
to which I discuss below). The above sentence in POL00006503 described just
one consideration as to the range of interests which ought to be borne in mind
as these strategies were developed. The Claimants were being funded by
Therium, a large litigation funder, and I presumed that their solicitors were acting
under a conditional fee agreement (as is ordinarily the case when a litigation
funder is involved). Naturally, the legal risks of the claim and the value of any
settlement pay out were highly important to them. These were legitimate
considerations to take into account. This was adversarial litigation with risks and
costs issues on both sides, and understanding the interests and aims of one’s
opponent is a very important aspect of litigation strategy. All this statement did
was identify what certain of these interests and aims were.
Page 205 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
360. At the October 2017 CMC, the Court adopted the approach that WBD/Counsel
had recommended to POL — namely to focus the first trial in the litigation on the
relationship between POL and its SPMs. A further CMC was listed for early 2018
to consider the scope of future trials. On advice from David Cavender QC,?*!
POL proposed that there be a Lead Cases trial to follow the CIT, the idea being
that the parties would identify 5-10 lead cases and have all the issues in those
cases determined in one trial. This would then create findings on issues such as
the reliability of Horizon, POL’s accounting practices and POL's training and
support, which could be applied by analogy to the other Claimants. I recall that
the Claimants initially supported this idea,?9? but Mr Justice Fraser preferred a
series of staged trials and ordered that the next trial be focused on the Horizon
system.
Advice regarding settlement strategy
361. We considered and advised on the question of settlement early on, and it was
revisited on various occasions throughout the proceedings.
362. On 8 July 2016, as part of our work on POL’s response to the Claimants' LOC,
WBD produced a paper on whether POL should engage in further mediation.
That paper drew out both the advantages and disadvantages of mediation. The
disadvantages included that the claims were poorly particularised and that
mediation was therefore unlikely to lead to a settlement. I was also concerned
that the claims had not yet been valued by the Claimants and that made it difficult
to advise on a potential settlement. Our main recommendation was that POL
291 POL00251957. See further below, §391.
282 POL00252386; WBON0001377; WBON0001378.
293 POL00006360.
Page 206 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
should politely refuse mediation at this stage but say they would keep the position
under review.
363. Settlement was revisited in a Steering Group paper dated 14 February 2017, but
for the same reasons WBD advised that settlement could not be considered until
the claim was valued and the group was closed to new Claimants. WBD therefore
recommended aiming for mediation in November 2017 after the first CMC (in
October 2017).2%
364. In November 2017, a Steering Group paper advised that “Post Office should
consider again whether there is merit in trying to settle this litigation. In particular,
there is an obvious window for a mediation in September / October 2018 to
explore the possibility of settlement before Trial 1 and in light of any risks flagged
by Counsel's advice”.29° The October 2017 CMC Order required POL and the
Claimants to explore the possibility of settlement. By a Steering Group paper
dated 6 December 2017, WBD identified two groups of Claimants with whom
early settlement might be possible and advised that settlement discussions be
commenced with Freeths in respect of these two groups. The notes (in blue
and red) of the views which different members expressed indicate in broad terms
that the Steering Group's view was that settlement discussions should not be
commenced at this stage without more considered discussion. Rodric Williams’
view was that settlement at this stage on the terms that WBD proposed ran the
risk of increasing the number of claims in the long term, consuming resources,
and weakening POL’s overall position with questionable return.2°”
294 POL00247209.
295 POL00139476
296 POL00251998.
287 POL00251998.
Page 207 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
365. The second CMC Order dated 8 February 2018 directed the parties to “use their
reasonable endeavours to attend mediation as soon as practicable after receipt
and consideration of the Judgment on the Common Issues to attempt to resolve
(or at least narrow) the dispute by way of mediation”.2°* At a meeting of the Board
Subcommittee on 15 May 2018, David Cavender QC and Tony Robinson QC
advised that “We should always keep the possibility of settlement under review.
But at the moment we do not see any other realistic option than to go ahead with
the Common Issues trial. There is then a mediation ordered to seek to settle the
matter and/or reduce the issues in light of the Common Issues judgement. There
is a very short time between that mediation and the Horizon trial.”2°° WBD
advised on concrete plans for mediation by a Steering Group paper dated 28
November 2018. This paper advised that settlement was the most likely outcome
and advised that POL write to Freeths on a without prejudice basis suggesting
that the parties start working on arranging mediation.5°
366. We wrote to Freeths on 7 December 2018*°' with our proposals on a mediation.
In light of the fact that we believed the Common Issues Judgment might be
handed down around the end of January 2019, we suggested a mediation in mid-
February 2019 might be viable.
367. We invited their response by 14 December 2018 but they had not responded by
then; we therefore chased for a response on 19 December 2018. Freeths then
responded on 21 December 2018. Their view was that a mediation after
Judgment had been handed down in the Horizon Issues Trial was more likely to
298 WBON0001230.
299 POL00006382.
300 POL00259669.
301 POL00265780.
Page 208 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
result in resolution or a significant narrowing of the issues than a mediation in
mid-February. They suggested a mediation mid to end June 2019.92
368. I remained of the view that it would be possible, and preferable, to hold a
mediation sooner rather than later. WBD therefore replied on 9 January 2019,
referencing the Order of 8 February 2018 and stating that in our view it was “quite
possible for the parties using their reasonable endeavours to hold a mediation
before the Horizon Issues Trial’. We explained that we believed a mediation in
February 2019 would set a foundation for future mediations and, at the least,
allow the parties to understand each other's position with more clarity. In view of
the Claimants reservations and with a view to engaging constructively with them,
we also suggested that the parties may want to appoint a mediator and seek their
views on when a mediation would most usefully be held.5°%
369. Freeths replied on 17 January 2019 to say that they did not believe the Court's
Order required the parties to mediate prior to the HIT. Their view was mediation
should take place after we had both the Common Issues and Horizon Issues
Judgments.°* In subsequent correspondence, the parties mutually selected
Charles Flint QC as a mediator but did not settle on a mediation date.
370. Ultimately, due to the Claimants being unwilling to mediate until after the HIT and
the Common Issues Judgment not being handed down until the HIT had already
started, there was no opportunity to mediate until later in 2019.
371. Abriefing on settlement was provided to Ben Foat, Rodric Williams and others in
May 2019.°° As set out in the briefing paper, the consistent view of the Steering
302 POL00260751.
303 POL00265783.
304 POL00262338.
305 POL00023690 and POL00275113.
Page 209 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Group and the Board Subcommittee to date had been that settlement could result
in a flood of claims and for this reason, POL’s preference had been to secure a
positive judgment on the Common Issues before opening up settlement
discussions in the expectation that this judgment (if positive) would deter future
potential claimants. Another obstacle to settlement continued to be that the
Claimants had not provided sufficient clarity as to the value of their claims.
Further, there remained some difficulty in respect of convicted Claimants. On the
advice of Cartwright King and Brian Altman QC, POL’s position from as early as
the Mediation Scheme was not to mediate or settle with convicted Claimants. In
this settlement briefing paper, WBD advised that fresh advice should be taken
from Cartwright King or Brian Altman QC on this issue.
372. Following this, POL took further steps to stand up a mediation with the Claimants.
HSF had been engaged by POL by this point, and from here onward took over
the lead in preparing for and conducting mediation. As explained above, a nine-
day mediation ultimately took place in December 2019, as a result of which the
proceedings were settled.
Relevance of the merits of defending the group litigation to POL’s review of
criminal convictions
373. With reference to Q58.7, I cannot give an informed view of whether the merits of
defending the group litigation affected POL’s review of criminal convictions or
informed its approach to post-conviction disclosure to convicted SPMs. I did not
advise POL on these matters, as my remit was limited to the civil litigation. As I
have highlighted in other parts of this statement, POL’s criminal law solicitors,
Cartwright King, supplemented from time to time with advice from Brian Altman
Page 210 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
QC, advised on criminal law matters including the safety of convictions and the
conduct of POL’s criminal law disclosure duties.
The Claimants’ Schedules of Information
374. Q70 of the Request asks me to specifically consider three documents:
374.1.First, my fourth witness statement in the group litigation, dated 9 October
2017 (POL00000444).
374.2.Second, a letter from Freeths to WBD dated 16 October 2017
(POL00041510), which responded to particular points in my fourth witness
statement.
374.3.Third, an email I wrote to Rodric Williams on 16 October 2017
(POL00041509).
375. With reference to Q70.1, Q70.3 and Q70.4, paragraph 26 of my fourth witness
statement commented on the poor quality of the SOls that had been prepared
on behalf of each Claimant, and observed that they appeared to have received
minimal input from Freeths. By way of context, the SOls were served in lieu of
individual particulars of claim and were, in effect, a summary of each Claimant's
claim. They were the only documents produced setting out the nature of each
individual Claimant's claim, and thus they were the sole source of information
within the proceedings as to precisely what each Claimant was seeking. They
were important because in their absence POL would not know (i) what each
Claimant was alleging, (ii) which claims were sufficiently similar so that lead
cases could be selected, and (iii) the amounts that were being claimed. Further,
a level of detail was required because the Claimants were alleging deceit against
POL, and their deceit claims were at this point unparticularised (i.e. the Claimants
Page 211 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
had not set out what they alleged POL had said, to whom, why it was false, and
with what consequences). The deceit/concealment point was critical, as it would
bear on whether the Claimants would be able to extend the otherwise applicable
limitation periods and many of the Claimants' claims would be time-barred
without an extension. Reflecting all these considerations, the SOls were required
to be signed by a statement of truth.
376. Freeths had opposed providing meaningful SOls. At the GLO Hearing, Senior
Master Fontaine ruled against them on this point and ordered the Claimants to
provide much more detailed SOls than Freeths had wanted to do (although, not
containing quite as much detail as POL had sought).
377. I was therefore disappointed when we eventually received SOls which were, in
my view, of a very poor quality. We sent a letter to Freeths dated 1 September
2017 in which we highlighted the extensive deficiencies in the SOls.° By way
of example:
377.1.Information pertaining to the Claimants’ allegations of deceit was largely
missing.
377.2.Several heads of loss claimed in the SOls seemed to have no actionable
basis and/or did not follow established legal principles. By way of example,
a high proportion of the Claimants (at least 65%) had claimed personal
injury. However, distress alone does not normally surpass the threshold
for bringing a personal injury claim; for that there needs to be a recognised
medical condition.
306 WBONO001194.
Page 212 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
377.3.There was little or no information about the nature of the problems each
Claimant claimed to have experienced with Horizon, for example
allegations of unusual behaviour or unexplained transactions and
accounting entries.
378. It was in part by reason of this second point that I referred to the Solicitor’s Code
of Conduct: Indicative Behaviour in paragraph 26 of my fourth witness statement,
which (so far as relevant) provided at the time that:
“demanding anything for yourself or on behalf of your client, that is not
legally recoverable, such as when you are instructed to collect a simple
debt, demanding from the debtor the cost of the letter of claim since it
cannot be said at that stage that such a cost is legally recoverable.”
379. I considered at the time that Freeths were close to this line; there were so many
inconsistencies in the SOls that it was hard to imagine that they had verified that
all of the claims contained therein were in fact recoverable as a matter of law.
That being said, I stopped short of accusing Freeths of misconduct in my fourth
witness statement as we could not be certain what work they had done on the
SOls. That is why the statement is framed in terms of what I anticipated or
expected, rather than an accusation of misconduct. I wished to make the point
that the quality of the SOls was far below what one would ordinarily expect to be
produced as a quasi-pleading in civil litigation, with the aim of trying to secure
improved SOls. I considered that it was proper to highlight relevant professional
standards which I believed should have been complied with in the preparation of
these documents, but I judged that it would not be right in the circumstances to
go further than this.
380. I therefore do not believe that I or WBD made an allegation of professional
misconduct against Freeths. We raised the serious deficiencies in the Claimants’
Page 213 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
SOls in firm terms to achieve a legitimate purpose, i.e. to obtain critical details
pertaining to the claims that had been made. I would not describe this as a
‘litigation tactic’ in a pejorative sense; it was a fair and proper response to
Freeths’ inadequate preparation of a key document in the proceedings.
381. At the CMC in October 2017, Mr Justice Fraser ordered that certain Claimants
amend parts of their SOls in relation to quantum.5°” Further, he required the
Claimants to obtain their medical records in support of their personal injury
claims, which was done due to doubts as to the credibility of those claims.
382. With reference to POL00041509 (my email to Mr Williams of 16 October 2017),
the remarks quoted in Q70.2 of the Request were a flippant comment on Freeths'
letter of the same date (POL00041510), which I attached to that email. The
context behind this comment was that at the time, I felt that Freeths were taking
small points for no substantive purpose. Freeths' letter followed an earlier one in
which they had complained that my fourth witness statement was filed 6 minutes
after the deadline.2° After that earlier letter, I had sent an email to the Counsel
Team (Tony Robinson QC and Owain Draper), saying:
“Another grumpy letter from Freeths about Parsons 4. I can't understand
this line of correspondence — it doesn't take them anywhere.’*"°
383. In fact, the response that I drafted to POL00041510 had the aim of minimising
wasted time on all sides by shutting this line of correspondence down, the proper
forum for the matters raised in my witness statement, and by Freeths, being the
CMC.3"! This was the appropriate and proportionate approach.
3°7 WBONO0001685, paragraph 21.
308 bid, paragraph 19; see also my fourth witness statement at paragraph 148.
309 WBONO0001216.
310 WBONO001215.
311 WBONO000191.
Page 214 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
384. For the avoidance of doubt, it was never my aim, nor (from what I observed) that
of POL’s legal team more broadly, to cause or encourage Freeths to waste time
on inter partes correspondence so that they spent less time on important matters.
Judgment No.1 [2017] EWHC 2844 (QB)
385. Q67 of the Request asks me to comment on the following statement of Mr Justice
Fraser in Bates & Others v Post Office Limited [2017] EWHC 2844 (QB)
("Judgment No. 1”), which was handed down in November 2017 following the
first CMC on 19 October 2017 (plus a short further hearing on 25 October to deal
with Tony Robinson QC’s availability for the CIT the following year):
“A fundamental change of attitude by the legal advisers involved in this
group litigation is required. A failure to heed this warning will result in
draconian costs orders’.
386. It is worth setting out the relevant paragraph in which this sentence is found in
full:
“Finally, litigation of any type, but particularly of this type, can only be
conducted in a cost-effective and efficient way if the parties co-operate
between themselves, are constructive, and conduct the case efficiently.
The parties have a duty to help the court to further the overriding
objective in CPR Part 1.3. The following have all occurred so far in this
group litigation: failing to respond to proposed directions for two months;
failing even to consider e-disclosure questionnaires; failing to lodge
required documents with the court; failing to lodge documents in good
time; refusing to disclose obviously relevant documents; resisting any
extension to the "cut-off" date for entries of new claimants on the Group
Register; and threatening pointless interlocutory skirmishes. On the
material before me, this has been more or less equally on both sides.
Such behaviour simply does not begin to qualify as either cost-effective,
efficient, or being in accordance with the over-riding objective. A
fundamental change of attitude by the legal advisers involved in this
group litigation is required. A failure to heed this warning will result in
draconian costs orders.”
Page 215 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
387. My immediate thoughts on reading Judgment No. 1 (then in draft) were reflected
in an email to Tony Robinson QC dated 8 November 2017. I explained that I
thought Mr Justice Fraser had got “the wrong end of the stick on many points”
(for example, the notion that we had opposed the GLO, see e.g. §350 above)
and that his “willingness to characterise points of disagreement [between the
parties] as unreasonable” made me nervous.*'? Tony Robinson QC responded:
“lL agree with all the wider points you make below. It is worrying that Fraser almost
seems to be one of those people who likes to think the worst of others, which
(entirely coincidentally) allows him to feel better about himself for sorting out their
deficiencies. I would like to comfort myself with the thought that his judgment
goes out of its way to lay equal blame on the claimants, but at this stage it would,
wouldn't it?”315
388. Tony Robinson QC went on to make the point that “Notwithstanding [Mr Justice
Fraser's] reference to pointless interlocutory skirmishes, if we have sensible
applications to make, we should make them, and make them promptly’.>"4 I
agreed with this observation and informed him that we now had instructions from
POL to make an application for security for costs. This was, in my view, a proper
application for POL to make, and I indicated that I would take certain
(constructive) steps first, namely (i) “callfing] Freeths to try to resolve this”, (ii)
writing direct to the litigation funder, and (iii) providing a draft application for
Freeths to comment on before issuing it.
312 WBONO0001217.
313 WBONO001217.
314 WBONO001217.
Page 216 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
389. As for the wider Steering Group’s discussion of Judgment No. 1, I cannot speak
to what conversations members of the group may have had among themselves,
but on 9 November 2017 Rodric Williams sent an email to Jane MacLeod,
Melanie Corfield, Mark Underwood, Mark Davies and Thomas Moran, making
the following points about the judgment:**5
“Mr Justice Fraser has used the judgment to reiterate the comments he
made at the 19 October 2017 CMC that the litigation needs to be
progressed in a more timely, cost-effective and proportionate manner
than it has to date, and that this will require greater cooperation between
the parties. He considers the failure of the parties to do so to date lies
‘more or less equally on both sides” (see para. 20)
[ud
Main Message
The tight timetable set for trial in November 2018 will not be departed
from, and the parties (through their legal advisors) will need to cooperate
to achieve this. Failure to do so “will result in draconian costs orders” (i.e.
the Court will order payment of substantial costs to the other side).
LJ]
What this Means — Longer Term
We must ensure that we not only cooperate with Freeths to promote the
expeditious resolution of the case (which we have been trying to do), but
that we are also seen to be doing so. Doing otherwise will irritate and
alienate Mr Justice Fraser, who will be presiding over the trial(s) in this
case. This must be kept firmly in mind as we plan and resource the next
12 months of this case.”
390. Following the judgment, on 12 December 2017, I sent an email to Freeths calling
for a reset in the correspondence and a “a better way of working between [the]
two firms”. As to correspondence, I outlined that since the CMC WBD had “sought
more than ever to avoid point-scoring in our letters” and adopt a constructive
315 POL00041527.
Page 217 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
tone. I set out a few points that would help me to run the litigation from the POL
side, i.e. fuller explanations/proposals from Freeths and as much information as
possible about the claims that were being advanced. I emphasised the need to
adopt a flexible, creative and bespoke approach to the group litigation, which
required “a good deal more collaboration than ordinary litigation if good progress
and efficiency [were] to be maintained.”>"®
391. As mentioned above (§333), David Cavender QC was instructed at this time to
lead on the Common Issues. At the outset, WBD asked him to take stock of the
litigation to date and to identify five things POL had ‘done well’, five things that
‘could have been done better’, and five things for us to ‘think about going
forwards’ (the “Five Things Document’).°"” In his Five Things Document, David
Cavender QC endorsed POL’s decision to accede to the making of a GLO, our
approach to the contractual issues and the setting up of the CIT as the first trial,
and our overall approach to the inter partes correspondence to date. In terms of
things we could have done better, he felt that POL “could have done more to
prevent the Claimants painting [it] as a party who was not co-operating properly
in the spirit of group litigation”, but he noted that (i) he did not feel this was a true
reflection of how POL had in fact engaged in the litigation to date, and (ii) we had
recently sent the above email to Freeths in the hope of ‘resetting’ the relationship
between the parties and paving the way for a more cooperative approach going
forwards. He said that we should now be looking to suggest “positive ways in
which the core of the bulk of the claims can be determined — rather than merely
seeking to respond/shoot down inappropriate ideas put forward by the
316 WBON0000329; see also above, §334.1.
317 POL00251957.
Page 218 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Claimants”, and recommended (as I have already mentioned, at §360) that we
should adopt a strategy of seeking to have a Lead Cases trial listed to follow the
CIT.
392. We therefore outlined our ideas for the future case management of the litigation
in a letter to Freeths dated 18 December 2017; we set out the Lead Cases
approach as recommended by David Cavender QC and stated, “We would
therefore welcome the opportunity to discuss the long term plan for this litigation
with you.... We are prepared to discuss all of the above at our meeting on 22
December 2017 but appreciate that there is a lot to consider in this letter... Our
client is not wedded to the proposals in this letter: they are just initial ideas to
hopefully encourage a constructive dialogue. We would welcome other ideas
from your clients’ .*"° I explained the reasoning behind the Lead Cases proposal
(as opposed to an approach of resolving issues on a topic-by-topic basis) as well
as the tone of this correspondence to Mark Underwood of POL on 17 December
2017, as follows:
“The letter will have a tone suggesting ways forward rather than making
a firm proposal and will ask Freeths for their ideas. We do not intend to
explain why the alternatives do not work (which was one of your
questions Tom) because we do not want to set a negative tone that
suggests we are blocking ideas or being difficult. If Freeths present an
idea that has merit, we should consider that in good faith. Our letter will
not therefore commit Post Office to a course of action, and will leave
scope to change direction if a better route opens up or we encounter
major resistance.
[ud
I hope this helps explain why we don't believe that there is an obvious
way for dealing with this litigation on a topic by topic basis. As said
above, if Freeths do come up with a solution to this, then we should give
318 POL00252386.
Page 219 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
it due consideration and our letter will be designed to draw them out on
this. "919
393. Thus in answer to Q67, and as the above points demonstrate, I and (to the best
of my knowledge) POL, considered the judgment carefully and took on board the
points Mr Justice Fraser made in an appropriate and measured way. We took
advice on our overall litigation strategy, which was broadly endorsed by Counsel,
and took positive steps to ensure a collaborative and constructive approach with
the Claimants going forwards.
L. THE GROUP LITIGATION - EARLY WORK
PART I — PRESERVATION OF DOCUMENTS, PRE-ACTION PROTOCOL DISCLOSURE,
PRIVILEGE AND SHARING OF INFORMATION (Q64 to Q65, Q68 to Q69, Q71 to Q72)
394. I return now to the start of the group litigation, and in this section answer a series
of questions the Inquiry has asked about particular aspects of the advice that
VWBD provided POL in its early stages. These questions are loosely centred
around the related themes of preservation of documents (Q64), early requests
for disclosure (Q68), advice on legal professional privilege (Q65, Q71 to Q72),
and other forms of information-sharing (Q65, Q67, Q71).
395. Broadly speaking, the events to which these questions relate all took place
between April 2016 (around the time when the Claimants’ Letter of Claim was
sent) and prior to the Claimants providing their draft GPOC in December of that
year. These events therefore overlap chronologically with those discussed in
Section M below, (concerning the early factual investigations undertaken by POL
319 WBONO000171.
Page 220 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
in the preparation of its LOR). In addition, it should be noted that in some respects
this section extends beyond 2016. For example, my answer to Q64 refers to
advice on preservation which WBD gave after 2016, and Q71 largely focuses on
matters that arose later in the litigation. I deal with that question here for
convenience, since it concerns advice I gave POL in connection with the sharing
of information about the group litigation with UK Government Investments
(‘UKGI’) and the then Department for Business, Energy and Industrial Strategy
(“BEIS”).
396. For the avoidance of doubt, this section does not deal with the Inquiry's questions
about my/WBD’s advice on disclosure generally, including its specific questions
about disclosure of the KEL, the Peaks database, and various reports produced
by Deloitte. These issues are dealt with in Section N below on disclosure.
(i) Document preservation (Q64)
397. I am asked to consider POL00041136, which is an email I sent to Rodric Williams
dated 20 April 2016. It contains an action list following a meeting attended by
Gavin Matthews and Elisa Lukas (WBD), Mr Williams and Mark Underwood
(POL), and myself on the subject of the group litigation. I have no recollection of
this meeting. Looking at when it took place, it must have been shortly after the
issue of the Claimants’ first Claim Form and in anticipation of receiving the LOC.
398. WBD undertook a considerable amount of work in the early stages of the group
litigation to assist POL to preserve relevant documents, and indeed throughout.
It was certainly, to my recollection, a long process that ran continuously during
the litigation, and one that evolved and expanded as the litigation took shape. I
believe that we had over 10,000,000 documents preserved in a data room by the
Page 221 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
end of the litigation. I note by way of background that there was some uncertainty
as to what needed to be preserved, especially in the early phases. Prior to our
receipt of the LOC, the claim had not yet been articulated for the purposes of
document preservation. Further, we did not have a full list of Claimants until the
GLO closed to new entrants in December 2017. Nevertheless, the overall tenor
of our advice to POL was that in the circumstances of the case and in light of the
issues raised, a robust approach needed to be taken to document preservation.
399. As I was supervising the firm’s work on document preservation and not directly
responsible for carrying it out, and given the volume and complexity of the work
required in this regard, I outline the main steps which WBD took in high-level
terms only:
399.1.First, WBD reiterated POL’s duty to preserve documents at the outset of
the litigation. Following the meeting referred to in POL00041136, Mr
Williams sent a litigation hold email on 20 April 2016 to all relevant staff at
POL outlining three document rules that must be followed, in these terms:
“In short, the three crucial document rules that must be followed
are:
(1) You must not destroy or delete any documents which may be
relevant to the claim. In particular, make sure that any automatic
deleting/archiving systems are suspended now until further
notice. If you have any question about whether a document is
relevant, please contact Legal Services and preserve the
document in the meantime;
(2) You must not amend any existing documents which may be
relevant to the claim. For example, do not make handwritten notes
on existing documents or try to change the content of a document;
and
(3) You must recognise that any documents that you create from
now on may have to be disclosed to the other side in the case. If
Page 222 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
in any doubt, think about whether you would be happy for the
email or document to be read out loud in court.’°2°
He attached a more detailed note on the subject dated December 2014,
which had been prepared by WBD at that time in response to a mooted
class action that never in fact materialised, together with a schedule of the
91 Claimants who were then listed on the Claim Form.%2"
399.2.Second, WBD undertook extensive fact-finding exercises with relevant
people at POL in order to find out how and where POL was holding
relevant documents. This included sending out questionnaires to
document custodians, carrying out interviews and follow-up exercises, and
engaging in discussions with the Company Secretary (regarding retention
policies), the IT team (regarding document creation, migration, storage,
extraction and deletion), and the Issues Resolutions, Support Services,
Agency Contracts and Contracts Advisors teams (regarding potentially
relevant documents and document sources).522 See, by way of example,
an email that Tom Porter (an Associate in my team) sent Dave King at POL
(who had assisted with the Mediation Scheme investigations) on 31 May
2016, which asked him and others to help WBD work out:
“What potentially relevant documents exist;
Where they are stored (and whether they are periodically backed
up);
Who is the stakeholder/controller for those documents;
Is there a retention policy that affects those documents (that may
result in them being lost unless otherwise preserved); and
320 WWBONO0000987.
321 WBON0000987; WBON0000988; POL0024 1034.
322 WBONO000151.
Page 223 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
What would we need to do to now protect and/or take copies of
those documents.’
My impression was that POL had very little understanding of how it held
documents; therefore, we were seeking to build up this picture essentially
from scratch. This culminated in the August 2016 Steering Group paper
outlining document preservation options to which I refer below.
399.3.Third, in August 2016 WBD prepared a paper for the Steering Group
setting out: POL’s duties to preserve documents; what steps had been
taken so far to identify relevant documents and repositories of documents;
what techniques were available to preserve material; the likely cost; and
our recommendations in view of all of these matters; and what was
understood about the claim to date. The paper advised:
“Post Office has a Court duty to take reasonable steps to preserve
any documents that may need to be later disclosed in the
litigation. "Document" means practically anything holding
information, including electronic documents like emails. What will
satisfy the duty to preserve documents will depend upon the
likelihood of documents being lost, how they may be lost and the
consequence on the litigation of losing a document.
[el
Steps to date
At the outset of the Group Action, Post Office Legal sent "litigation
hold notices" to key parts of the business asking them not to
destroy relevant documents. Since then BD has liaised with
various teams at Post Office regarding potentially relevant
documents and document sources. Through these investigations,
we have developed an understanding of document storage,
retention and deletion across the business, as well as better
understanding the current IT projects that may impact on
document preservation. Please see the Document Locations
Table attached to this paper for details of the locations in which
323 WBON0001002; see further WBONO001015.
Page 224 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
documents are held. These investigations have led to the
development of the "Preservation Options" attached to this paper.
[ood
Our view is that some form of limited forensic imaging of
information is required — either of documents held by key
custodians (Option 3) or by undertaking a deeper review to
identify more relevant locations of documents (Option 4).”924
The paper highlighted the advantages of taking robust as opposed to
minimal steps to preserve documents, including that: “The nature of the
claims in this matter, particularly the fraud and concealment issues, means
that preservation is a relatively high risk issue in this case. Losing key
documents where there are allegations of concealment would weigh
against Post Office in Court and would be presented by Freeths as yet
another form of concealment. This militates towards Post Office taking a
more stringent approach to document preservation”. And that: “Doing
nothing risks falling foul of the Court duty to preserve relevant documents.
Aside from the legal consequences, this would present very badly through
a public / media lens”. The Steering Group's decision was in line with our
recommended approach, adopting our Option 4 as POL’s preferred
approach to document preservation.32°
399.4.Fourth, we advised POL on the importance of ensuring that third parties
(notably Fujitsu) also took steps to preserve relevant documents.
POL00041136 refers to a “letter to FJ re document preservation’. I believe
that this was drafted by WBD and provided to Mr Williams at an early
324 POL00139297.
325 POL00139309 (action from 22 August 2016 meeting).
Page 225 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
stage, on 26 April 2016.26 An email sent by Tom Porter on 6 May 2016
recorded our understanding that Mr Williams was “writing to Fujitsu to put
them on notice of the claim, and ask them to ensure that potentially
relevant documents are now retained” and advising that “We should [also]
give some thought to whether we need to send similar notices to other
third party providers at this stage”.5”” It appears from subsequent emails
in November 2016 that the letter to Fujitsu may not ultimately have been
sent. On 15 November 2016, Elisa Lukas reported to me that Mr Williams
had told her that he had “not informed [Royal Mail] or Fujitsu of the need
to preserve documents as he does not consider their documents to be in
his possession or control and it will be costly to [POL]’.°28 I noted the need
to look into the control issue but thought that POL needed to send litigation
hold notices to POL and Royal Mail regardless. WBD then drafted an email
to Fujitsu and Royal Mail advising them as follows:°2°
“As you may be aware, a group of former and current
postmasters, branch assistants and Crown Office employees
have brought a legal claim against Post Office in relation to
Horizon. Their claim is very broad, alleging failings in Horizon as
well as Post Office's training and support (the Action). A copy of
the Claim Form is attached.
In light of this proposed litigation, please can you ensure that all
documents that you hold on behalf of Post Office and which are,
or may be, relevant to the Action are preserved. Please ensure
that this includes any electronic documents (and associated
metadata) which would otherwise be deleted in accordance with
your document retention policy or in the ordinary course of
business.”
326 WBON0000982; WBONO0000981.
327 WBONO0000992.
328 WBON0000154.
329 POL00041378.
Page 226 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
399.5.Fifth, in August 2017, WBD prepared a further paper for the Steering
Group on the subject of document preservation. This was triggered by the
fact that a considerable number of additional Claimants had by then joined
the litigation. The paper outlined the approach taken to date as follows:
“1, BACKGROUND
1.1 Post Office has a Court duty to take reasonable steps to preserve
any documents that may need to be later disclosed in the litigation.
At the Steering Group meeting on 22 August 2016, Post Office
decided to take a proportionate approach to this duty. ...
1.2 It was decided that a list of key individuals across the business
who might hold relevant documents would be produced, and then
establish what documents they held and how. Forensic copies
would be taken of relevant electronic documents and scanned
copies taken of hard copy files. Although not all relevant
documents would be preserved because of the targeted nature of
the exercise, it would demonstrate a genuine attempt to preserve
documents.
1.3. It was recognised that that the preservation exercise would need
to be refreshed if/when further Claimants issued a claim against
Post Office ... and as the litigation progressed.
2. DOCUMENTS ALREADY PRESERVED
2.1 The preservation exercise had a dual purpose: it was to preserve
documents and also to provide information to support the Case
Review exercise. Within the original 198 Claimants, 88 were part
of the mediation scheme and so the Case Review was limited to
the other 110 cases.
2.2 This has led to a tiered capture of documents:
2.2.1 Litigation hold notices have been sent to key parts of the
business. This covers a wide range of documents but
only provides a low level of assurance that documents
will not be destroyed.
2.2.2 For some categories of documents, we have extracted all
Post Office data in relation to all subpostmasters (not just
Claimants).
330 POL00006436.
Page 227 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
2.2.3 For some categories of documents, we have only
captured information relating to the 198 original cases.
2.2.4 For some categories of documents, we have only
captured information relating to the 110 cases subject to
the Case Review.
2.3 In general, we have narrowed the capture of documents when
dealing with paper records as these require significantly more time
and cost to locate and scan into a data room.
2.4 The preserved documents are being hosted in a data room that
currently holds 599,004 documents. Further information on the
documents that have been preserved already can be found in
Schedules 2 and 3.
2.5 The focus of work so far has been around preserving documents
relevant to individual Claimants. We have not yet preserved
documents relevant to generic issues. For example, we have
captured the debt team files on individual Claimants, but we have
not scoped and preserved general documents and policies about
debt collection practices. This is because "generic" documents
are much more difficult to identify, locate and retrieve in a cost
effective way.”
The paper went on to note that, in light of there now being 324 additional
Claimants, a decision needed to be made as to the extension of the
document preservation exercise. We identified that consideration should
now be given to extending the exercise to: (i) documents relating to the
new 324 Claimants;**" and (ii) certain areas of generic documentation. We
recommended:3%2
“As a minimum, Post Office should extend the document
preservation exercise that has already been carried out to the 324
new Claimants.
We would strongly recommend preserving the entirety of POL
SAP when it is taken offline later this year. Currently, only certain
331 POL00006436.
382 I.e. by the issue of the Claimants’ second Claim Form on 26 July 2017.
Page 228 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
reports have been run from it and it contains a vast amount of
potentially key financial information on the Claimants and their
branches.
We would also recommend beginning to capture key generic
documents that are likely to be sought in disclosure by the
Claimants, to the extent that this could be done cost-effectively.”
The paper noted that “/n the Decision Paper of 22 August 2016, [WBD] set
out the advantages of preserving documents. We continue to believe that
these advantages justify the above costs’. I do not specifically recall
whether the Steering Group adopted the recommended course, but I recall
that POL generally accepted the advice given on issues of this kind.
399.6.Sixth, as issues for disclosure and classes of document to be disclosed
were ordered during the CMCs from October 2017 to June 2018, the scope
of the preservation exercise was revisited, and where needed expanded,
to ensure that those issues and classes were covered.
399.7.Seventh, at various stages WBD provided targeted advice about specific
document preservation issues. For example, by a Steering Group paper
dated 24 May 2017, WBD “strongly recommended” that POL continue to
pay for a hold on data which Fujitsu was preserving to prevent it from being
deleted.5°° Formal advice to the Steering Group on other strategic issues
relating to document preservation was given in January 2017,°%* May
383 POL00139383.
334 WBON0001686.
Page 229 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
2017,355 September 2017,3% March 2019,397 July 2019,°98 and November
2019.389
399.8.Eighth, beyond the formal Steering Group papers, I recall my team
regularly discussing discrete document preservation issues with POL. I
cannot now recall or cite all the issues that came up, but as a few
examples, I recall: discussions about copying and preserving laptops of
members of staff who were leaving POL; members of my team having to
visit POL's Chesterfield office in order to determine whether physical files
in that office needed to be preserved; and many discussions about
preserving and accessing records held at the Postal Museum, which POL
used as a repository for hardcopy documents.
(ii) Data Subject Access Request (Q65)
400. By Q65, I am asked to comment on POL00041163, which is an email chain
relating to a Data Subject Access Request (“DSAR’”) by Katherine McAlerney
dated 27 April 2016. I am in copy from the second email. This email chain shows
that in relation to Ms McAlerney, the POL team responsible for dealing with
DSARs (Kerry Moodie and Kim Thomson) had previously provided the
information she had requested and which they could locate. The POL Security
Team located some (very limited) further information, which Kerry forwarded to
me for comment on 1 June 2016. That information was a single row extracted
from a spreadsheet that no one from the POL Security Team (past or present)
335 POL00006405.
336 POL00006470.
337 POL00269447.
388 POL00278526; POL00139652; POL00139650.
398 POL00288913.
Page 230 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
could recall preparing (nor the reasons for preparing it).*4° Prior to this, Paul
Loraine and Rodric Williams had produced a draft response to Ms McAlerney’s
DSAR that explained POL did not propose to take any further action because this
was an improper use of the subject access regime to obtain documents outside
of the disclosure process in the group litigation.°41
401. I responded to Kerry Moodie in the following terms the same day:
“The new information found by POL security makes reference to this
case being handled by "legal". On that basis, we can treat the
information (which is in any event very limited and inconsequential) as
being privileged and therefore not disclosable.
The current draft of the letter is therefore good to go.
If anyone objects to this approach, please can you let Kerry know by
3pm tomorrow otherwise — Kerry please can you send the letter in
tomorrow's post.”
402. By way of context, it is important to understand that a DSAR operates quite
differently from a disclosure exercise. In particular:
402.1.A DSAR is a request for personal data; it is not a request for
documentation. Much of the information contained within this spreadsheet
row was personal data that was already known to Ms McAlerney (such as
her branch name, FAD code, her name, the loss amount, and that her
solicitor had responded to an intimated civil claim by POL). The only new
information was the part of one sentence that said that POL had “escalated
to Legal to pursue”.
340 WWBONO001013.
341 WBONO0001001.
Page 231 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
402.2.The data controller is only obliged to take reasonable steps to find personal
data and need not go to a disproportionate effort. The case law at this time
spoke in terms of a reasonable and proportionate search, such that where
there were face value grounds to assert privilege, there was no obligation
to extensively examine the basis for asserting privilege.*42
403. I do not recall this particular email or this issue. As indicated by the brevity of my
email, I would have made this type of judgment call rapidly, given that in the
DSAR context granular analysis of whether privilege applied was not required. I
do not believe that I was referring to the whole document as privileged, but rather
the particular sentence highlighted above, that the matter was being “escalated
to Legal to pursue”. Bearing in mind the lack of available background information
to contextualise to this document, and the references to ‘civil charges’ and the
fact that Ms McAlerney had appointed a solicitor, I would likely have considered
that there was a fair argument that the sentence was either privileged information
itself or a reference to other privileged material, namely POL's confidential
decision to consider pursuing a legal claim and to seek legal advice thereon.
(iii) Pre-action disclosure of POL’s internal investigation guidelines (Q68)
404. I am asked to specifically consider POL00038852, which is an email from Amy
Prime to Rodric Williams (I am in copy) dated 5 October 2016.*4% I do not recall
this email, but I have reviewed it along with other relevant emails from the time
in order to answer Q68 of the Inquiry’s Request.
342 See, by way of example, Dawson-Damer [2015] EWHC 2366 (Ch) at [34]-[37]. This decision
was overturned on appeal in 2017, but represented the law as I understood it in June 2016.
343 Incorrectly dated 10 May 2016 at Q68 of the Request.
Page 232 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
405. Amy’s email concerns a request Freeth’s had made in its LOC dated 28 April
2016 for POL to disclose its “/nvestigation guidelines since 1998, including any
revisions to date.”*“4 By way of context, this was one of 32 requests for disclosure
made in the LOC.
406. In our LOR dated 28 July 2016, we said the following in response to Freeth’s
request:
“We are currently reviewing this request and will update you in due
course. We understand that these guidelines will have evolved during
the period in dispute. Further, providing historic documents would
require a full disclosure exercise. This is neither reasonable nor
proportionate at this time.’°45
407. It appears that the “investigation guidelines” that we had available at this time
were a document relating the conduct of criminal investigations, being the
version in force from August 2013.°4° These guidelines were unlikely to be
relevant to the Claimants’ prosecutions given that POL had largely stopped
prosecuting SPMs for accounting shortfalls around that time. The likelihood
therefore was that only earlier versions of those guidelines would have been in
force at the times of the Claimants’ prosecutions. Consequently, in order to
comply with Freeths’ request, we would have had to conduct further searches to
piece together all historic versions of these documents dating back to 1998. We
would also have had to search for other forms of investigation guidelines because
Freeths' request was not just limited to criminal investigations. I was aware by
this time that POL undertook several different forms of investigation through
different departments (e.g. the teams that investigated disputed transactions
344 POL00241140.
345 POL00110507.
346 WBONO000466.
Page 233 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
corrections) and so any search for investigation guidelines would have been a
substantial exercise.
408. From my emails, in the immediate run-up to serving the LOR I had liaised with
Brian Altman QC on the subject of the criminal investigation guidelines, and he
told me that documents of this kind were not normally privileged from disclosure
in criminal proceedings**” (as I had mistakenly come to think based on earlier
advice given in a different context by Cartwright King).*4® As my exchange with
Brian Altman QC shows, I did not have any particular concerns about disclosing
the 2013 investigation guidelines, bar the possibility that doing so might lead to
the loss of some form of privilege. Indeed, I explicitly said that I considered their
content “pretty benign” .*49
409. Nonetheless, it was not necessary for POL to disclose the 2013 investigation
guidelines when sending the LOR, nor was POL required to conduct a wider
search for all previous versions of the guidance that might possibly have applied
at the time of the Claimants’ prosecutions. At this early stage, disclosure had not
been ordered and the parties were operating within the framework of the Practice
Direction on Pre-Action Conduct, which only required a party to take reasonable
and proportionate steps to identify, narrow or resolve issues in dispute and to
disclose key documents. The 2013 guidelines we had were not likely to be
relevant to the Claimants and were not necessary in order for the Claimants to
formulate their case, and it would not have been reasonable nor proportionate to
search for earlier versions at the pre-action stage.
347 WBON0000443.
348 See above, §192.
349 WBON0000443.
Page 234 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
410. It appears that WBD returned to the issue of the investigation guidelines in
September and October 2016. I highlight that at that stage, the scope of POL’s
duty of disclosure was no greater than it had been in July. My email records
indicate that Amy Prime obtained updated versions of the guidelines from POL
in September 2016 (i.e. the 2016 guidelines, which had replaced the 2013
iteration earlier that year) but was instructed by Jane MacLeod that POL did not
wish to allow them to be disclosed given that they could not, almost by definition,
be relevant to any of the Claimants’ prosecutions.3°°
411. Against that background, reviewing POL00038852 I surmise that Amy
considered that POL ought to maintain its position that the 2013 guidelines
should not be disclosed at this stage and that she similarly considered that the
2016 guidelines should not be disclosed. I cannot recall what if any discussions
I had with her about this at the time and cannot expand upon her reasoning,
whether as expressed in POL00038852 or otherwise. However, reviewing that
email now, I do not think that anything in it alters the fact that there were legitimate
reasons for declining to disclose the guidelines at that stage. As Jane MacLeod
had pointed out, the 2016 guidelines could not have had any bearing on the
convicted Claimants’ claims and the 2013 guidelines were very unlikely to do so
(even allowing for the fact that a number of Claimants had been added to the
Claim Form by way of pre-service amendment in July 2016).
412. With the foregoing in mind, I refer to the email at POL00038852 and in particular
the final paragraph of that email, which I am asked to consider.
360 WBON0000464.
Page 235 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
413. Regrettably, this email is worded very poorly. Whilst, as I have said, I do not recall
this email, my firm’s records show that Amy had sent a draft for my approval
earlier that day which did not contain this final paragraph.**" I responded to her
adding it into her draft, though my purpose in doing so appears to have simply
been to make clear what action we required from POL on this point, rather than
to consider or build upon the substance of her email.*? Though ill-expressed,
having reviewed the relevant emails from around this time I consider that
POL00038852 and the final paragraph in particular does not reflect the true
position, as there were in fact substantive legitimate reasons for resisting
disclosure of the investigation guidelines at this early stage. My email should
have been better expressed to make that clear at the time.
414. My firm’s searches suggest that POL were never later ordered to disclose the
2013 or 2016 investigation guidelines.
(iv) Reference to bugs in briefing note to Leading Counsel (Q69)
415. I am also asked to specifically consider POL00022636, which is an email I sent
Rodric Williams and Jane MacLeod on 18 May 2016, attaching a briefing note
(POL00156685). The background to that email and the production of the briefing
note was because POL was in the course of selecting its Leading Counsel and
was due to interview Jeffrey Onions QC and Tony Robinson QC. The briefing
note was to introduce them both to the background of the matter before the
interviews were held. The purpose of the briefing note is set out in paragraph 4:
351 WBONO0000465.
362 WBON0000467.
Page 236 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
“This [sic] purpose of this briefing note is to summarise the legal and
factual context of this dispute by way of a high-level briefing ahead of a
meeting between POL and prospective Counsel. This note has been
prepared from a compilation of other documents and any views
expressed below will need to be tested and verified once Counsel
is properly instructed and full documentation made available.”
(emphasis added)
416. The note lifted heavily from the Swift Review as this already set out a good
description of the background facts and the issues under consideration, and
there was no need to re-invent the wheel for the briefing to Counsel.°** Large
parts of the content of the Swift Review were copied, word-for-word, into it. The
sentence that is quoted at paragraph 95 (“There is nothing to suggest that these
specific bugs identified have been the cause of wider loss to SPMRs in the
Scheme cases or otherwise ...”) features word-for-word in the Swift Review (see
paragraph 120 thereof).
(v) Conference with Tony Robinson QC on 9 June 2016 and advice on preserving
privilege in the implementation of the Swift Review (Q72)
The Swift Review
417. Tony Robinson QC was instructed on behalf of POL and on 9 June 2016. I
attended an initial conference with him at POL’s offices at Finsbury Dials,
together with POL’s in-house legal team. Prior to that, I had attended a pre-
conference meeting with him on the morning of 7 June 2016 (accompanied by
Tom Porter, an associate in my team) to address any initial questions that he had
arising from his instructions.°4
383 WWBON0000993 and WBON0000179.
364 WBON0000157.
Page 237 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
418. As I have mentioned previously, at this time there was a live issue as to whether
documents generated by Tim Parker's proposals to implement the Swift Review
recommendations would attract legal professional privilege (see above, §§305-
308). In particular, POL was concerned that Deloitte’s work product would not be
privileged if they were instructed directly by Tim Parker to carry out the
investigations contemplated by Jonathan Swift QC’s IT-related recommendations
(being Recommendations (3), (4), (5) and (8)).
419. I do not recall the specifics of the discussion about the implementation of the
Swift Review at the conference on 9 June 2016. However, the documents I have
reviewed for the purposes of drafting this statement (including but not limited to
POL00006601) confirm that we sought Tony Robinson QC’s view on how POL
should proceed. The notes of the pre-conference meeting I had with Tony
Robinson QC on 7 June 2016 include a bullet point: “/djo all the Swift actions
now and thoroughly’ beneath the heading ‘key thoughts’ (presumably, those of
Tony Robinson QC).35>
420. My email records show that I had a more detailed exchange with Tony Robinson
QC about this issue on 8 June 2016, prior to our conference the following day.
During that exchange I made the following points:9°¢
420.1.First, the Steering Group had expressed concern about Tim Parker
continuing with his proposed implementation of the Swift Review
recommendations, however Mr Parker felt that he had “made a
commitment to Baroness Neville Rolfe (Minister at BIS) to follow through
385 WWBONO000157.
366 POL00242402.
Page 238 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
on the JSQC's recommendations unless he [was] presented with a
persuasive case not to do so”.
420.2.Second, the principal outstanding recommendations were that: (i) POL
should carry out an investigation into the issue of remote access to Horizon
data (which Tim Parker intended would be done by Deloitte); (ii) POL
should carry out an investigation into unmatched balances on POL’s
general suspense account (which it was again intended would be done by
Deloitte); and (iii) POL should review those cases where theft and false
accounting were charged simultaneously to establish whether there was
sufficient evidence to mount the theft charge (which review was already
being undertaken by Brian Altman QC).
420.3.Third, all three recommendations overlapped with issues in the group
litigation, and there were therefore three obvious reasons why Mr Parker
should not commission the relevant reviews to be conducted on his behalf:
(i) It would be necessary for the three points raised by Jonathan Swift QC
to be investigated in the course of the litigation. Running two parallel
sets of investigations would be costly and could cause difficulties if they
reached differing conclusions. Further, carrying out the investigations
in the context of the litigation would likely have the advantage of speed
and the conclusions reached being robustly tested in court.
(ii) If the investigations were conducted by Tim Parker there was a greater
risk that this work would not be privileged (since the investigations
would arguably not be conducted for the purposes of litigation but for
some other purpose).
Page 239 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(iii) The risk identified at (ii) above was likely to be compounded by the fact
that Mr Parker would wish to provide the outcome of the investigations
to the Government, which I thought might lead to the loss of any
privilege that might otherwise have been asserted.
420.4.Fourth, therefore, whilst there was an element of “political background” (in
that I was informed that Tim Parker had made a political commitment to
see the Swift Review recommendations through), POL’s interests in
defending itself in the litigation meant that a different approach was
appropriate and justified.
420.5.Fifth, if Tim Parker’s review was to cease, POL would have to reckon with
the risk that the Swift Review recommendations might not ultimately be
achieved (or be fully achieved) through the litigation: “the work is either
required for the litigation or it is not. We can't artificially squeeze work
under the litigation umbrella just to cover off a political issue (or at least
that is my view anyway)”.
420.6.Sixth, however, in all likelinood the investigations which would be needed
to aid the litigation would be “largely duplicat[ive of] what TP would have
been doing’. As such the proposal was to complete substantially the same
work but for the purposes of contesting the litigation.
420.7.Seventh, assuming Deloitte were instructed to proceed with one or both
of the remote access and suspense account investigations, my preference
would be to instruct a different expert as our witness in the litigation. This
was because Deloitte's previous close involvement might result in their
earlier instructions (which I had not seen) becoming disclosable. It might
Page 240 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
also call into question their independence for the purposes of CPR 35; in
this regard, Tony Robinson QC had commented that if Deloitte were to be
instructed as our expert witness, this could impact upon how closely we
were able to work with them.
421. The views Tony Robinson QC expressed in the course of this email exchange
were substantially the same as mine.°*” He agreed that he was “concerned that
the client should protect its interests as a defendant to this substantial piece of
litigation”, in relation to which he thought the “overriding [consideration was] the
privilege point’. At the same time, he strongly agreed with the approach of
subsuming the investigations recommended by Jonathan Swift QC into the group
litigation workstreams: “From a pure litigation perspective, these investigations
are highly desirable — the less evidence we have to rebut the suggestion that
remote data tampering at our/Fujitsu’s end could be responsible for inflicting any
false losses on any claimants, the more awkward our position is on this difficult
point’ (emphasis in original).
422. Although, as I have said, I cannot recall the specifics of the discussion about the
Swift Review at the conference the following day, it is evident from the letter
which WBD subsequently provided to POL that: (i) “Mr Robinson was asked to
advise on, amongst other matters, whether Mr Parker should continue his review
and/or implement Mr Swift's recommendations’; (ii) his “very strong advice” was
that the review should “cease immediately’ (with earlier drafts of the letter
speaking to the “material risk that Mr Parker’s review, and particularly the
implementation of Mr Swift's recommendations by Mr Parker, would not be
387 POL00242402.
Page 241 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
covered by privilege”);>** though (iii) he recommended that POL should still
implement the relevant recommendations of the Swift Review, albeit with that
work being “instructed and overseen exclusively by POL’s legal team (or others
instructed by POL’s legal team) so as to maximise POL’s prospect of asserting
privilege” (POL00006601).
423. Gavin Matthews took the lead on drafting this letter with some input from Rodric
Williams, but he did circulate it to me for my approval, as is shown by
POL00041242.
424. I believe the email exchange of 8 June 2016 fairly reflects my views at the time
on the question as to whether Mr Parker should proceed with his planned method
of implementing the Swift Review recommendations (i.e. under the auspices of
the Chairman's Review). Like Tony Robinson QC, I thought it highly desirable
that the investigations suggested by Jonathan Swift QC should still be carried
out, but as part of the ongoing work on the group litigation so as not to lead to
duplication of work, potential inconsistency of results, and (crucially) loss of
privilege in the product of those investigations.
425. I do not recall giving advice directly to Tim Parker or any other representative of
POL on whether to provide the Swift Review to the Board, UKGI and/or the
Government. As I set out below, my (limited) involvement in advising POL on
what information to share with UKGI and BEIS came later and was concerned
with the provision of privileged documents arising out of the group litigation to
those bodies. The searches my firm has carried out in this regard support my
recollection.
358 See Gavin Matthews’ first draft of the letter, dated 16 June 2016: POL00242578.
Page 242 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Advice on merits and strategy
426. I briefly summarise here the other advice that Tony Robinson QC gave during
the conference on 9 June 2016 (cf. Q72.1). As my email of 24 May 2016
instructing him shows, and as I have explained above, the primary concern at
this early stage of the litigation was whether POL should consent to the making
of a GLO. POL also wanted to understand which (if any) of the terms which
Claimants sought to the imply into the SPM contract should be admitted by
POL.®*? I recall that these were the focal points of the conference on 9 June 2016,
although I think POL was only looking for Tony Robinson QC to give his high-
level thoughts on these matters to get a sense of our general direction of travel;
there was no expectation that he would express any definitive views on the merits
at this stage.
427. I have some limited recollection of the views Tony Robinson QC expressed about
these points. In sum, I recall that he thought:
427.1.POL should accede to the making of a GLO in principle.
427.2.POL should not admit any of the Claimants' implied terms, but should
admit two other implied terms (namely, reasonable or necessary
cooperation and ‘Stirling v Maitland’ terms).
427.3.Remote access was, in his view, a major issue. That was because (i) POL’s
previous misstatements on that topic damaged its credibility, and (ii) it ran
the risk of extending the normal six-year limitation period.
369 WBON0000995.
Page 243 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
428. It is this final point that substantially increased the importance of the remote
access issue once the litigation began. Up to this time, there had been no
substantiated example of remote access being the cause of shortfalls in branch
accounts. With the focus of the Mediation Scheme being on establishing the
cause of shortfalls, there were many more lines of inquiry (other than remote
access) being raised by SPMs and Second Sight that were more likely to be
probative of what had happened in branches (e.g. accounting for scratch cards,
cash remittance processes, etc.). Also, limitation had not been in issue in the
Mediation Scheme because POL had elected to open the scheme up to SPMs
whose claims might otherwise be time-barred. Once the litigation began, it was
expected that many (perhaps even a majority of) cases would be outside the
ordinary limitation periods and limitation defences would be raised by POL for
the first time. So, the issue of remote access was, in my mind, of much greater
significance once the litigation began because of its role in potentially extending
limitation periods.
429. Beyond this, I do not recall that Tony Robinson QC provided much further advice
on general litigation strategy at the 9 June 2016 conference. At this time, he was
still getting up to speed with the case (he had only been instructed approximately
two weeks’ prior),°®° and we were still working our way through the lengthy LOC.
He therefore only proposed to express his preliminary views.
(vi) Information-sharing with UKGI and BEIS (Q71)
Overview
360 See by way of further background to Tony Robinson QC's early reading in: POL00140216.
Page 244 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
430. Q71 asks me to consider POL00041770, and to describe any advice I gave “on
the issue of POL sharing information with UKGI/BEIS and any ways in which the
same could be limited’. Other than one insubstantial occasion I do not recall
having any direct interaction with UKGI or BEIS about the group litigation.°'
These relationships were managed in-house by POL, and it was not part of my
role to advise POL on its strategy for dealing with UKGI and BEIS or its wider
relationship with those bodies. My overall impression was that I was only aware
of a fraction of communications between POL and UKGI / BEIS.
431. The nature of my involvement was that I advised, from time to time, on specific
issues concerning the provision of information or documents relating to the group
litigation to UKGI and BEIS. Invariably, this constituted advice on whether the
sharing of a document, or of information, would result in a loss of privilege or
confidentiality in the same. My focus was therefore not on limiting what POL
exchanged with UKGI and BEIS, but on ensuring that proper controls were in
place to prevent privilege being waived and to protect confidential and sensitive
information. In my view, this was a fair concern on POL’s part.
432. To the best of my recollection (and searches of my firm’s file appear to confirm
this), my involvement covered the following matters:
433. First, I advised on an information-sharing protocol between UKGI, the Secretary
of State for Business, Energy and Industrial Strategy and POL (the “UKGI/BEIS
361 In early 2018, after Tom Cooper was appointed to the POL Board, he was briefed on the
litigation by Jane MacLeod. I attended that meeting and there was a junior person from UKGI
in attendance. I can also see from my calendar that Richard Watson of UKGI was listed as an
attendee for a meeting on 11 April 2019 — I do not recall whether he attended that meeting or
not. Searches of my email records show that in May 2019 I sent case summaries relating to the
six Lead Claimants in the CIT to Tom Cooper, with UKGI staff in copy. This is the only email I
have been able to locate directly between me and UKGI: WBONO0000160.
Page 245 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Protocol”) that was drafted by POL, as I set out further below. This is the topic
to which the email chain at POL00041770 relates. The essential purpose of the
protocol was to regulate the provision of information about the group litigation by
POL to UKGI and the Secretary of State; to set out what would be shared, how,
what UKGI and the Secretary of State could do with that material; and to ensure
that documents provided under the protocol which were privileged remained so
by establishing unequivocally that POL, UKGI and the Secretary of State shared
a common interest in the litigation.
434. Second, I was sometimes asked to advise on the day-to-day application of the
UKGI/BEIS Protocol. For example, I confirmed to Rodric Williams that copies of
the expert reports for the Horizon Issues Trial could be shared with UKGI
pursuant to the protocol.*® In another instance, I made changes to a briefing
note which POL proposed to send to UKGI on the impacts of the Common Issues
Judgment — reformatting it as a note of advice to POL — to ensure that it would
attract privilege and retain it under the UKGI/BEIS Protocol (and this meant that
POL could be freer in what information was included in the note).°*° More
generally, I was occasionally made aware of briefings to be given to UKGI and
the Secretary of State under the Protocol and was sometimes asked for my
comments,°** but was not normally involved in drafting them.°® In particular
instances where WBD had the best handle on the underlying facts and matters,
POL would occasionally ask WBD to draft material that POL intended to provide
362 WBON0000648.
363 WBON0000662; cf. POL00023809 which concerned a similar briefing to BEIS, albeit one
prepared by a non-lawyer which would therefore not attract legal advice privilege.
364 See, by way of example, in relation to one of those briefings, which was forwarded to me for
comment on 8 June 2018, I recommended that additional technical controls were put in place
to avoid some of the risks of inappropriate onward use, e.g. conversion to PDF, the addition of
password protection, and so on: POL00024241.
365 See, by way of example, POL00041825.
Page 246 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
to UKGI under the Protocol. For example, I prepared a draft email to UKGI
providing an update on POL’s application for security for costs*®* and WBD also
worked on a draft paper for a litigation briefing that UKGI attended on the topic
of the appeal of the CITJ on 11 April 2019 (though it is not clear from my firm's
records who in fact sent the final version to Jane MacLeod).°6”
435. Third, POL would also sometimes seek WBD’s input on factual matters to be
relayed to UKGI or BEIS where we were closest to the material — for example,
information on timetabling®®* and the costs of the litigation.2®
436. Fourth, we occasionally advised on the sharing of other information or
documents not covered by the UKGI/BEIS Protocol. For example, I advised that
the draft Common Issues Judgment could not be sent to UKGI whilst under
embargo (notwithstanding the Protocol).9”° This was later also applied to the
Horizon Issues Judgment.$”!
437. Fifth, matters would sometimes come to my attention that I would (on rare
occasions) flag with POL’s in-house legal team as being of possible interest to
various stakeholders, including UKGI and BEIS. For example, I queried on 31
May 2018 whether UKGI should be informed of the Claimants’ first aggregated
claim valuation of £80 to £90 million, which featured in their skeleton for the CMC
on 5 June 2018. Jane MacLeod responded in the affirmative and prepared a draft
note to that end.3”2
366 WWBONO0001306.
387 WBONO000705.
368 WWBON0001249, WBON0001417, WBON0000691 and POL00023301.
368 WBONO0001643.
370 WWBONO0000641 and WBON0000647.
371 WBONO0000719.
372 WBON0001248.
Page 247 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Drafting of the UKGI/BEIS Protocol
438. POL00041770 is an email chain dated March to May 2018 which records (in part)
the negotiation of the UKGI/BEIS Protocol. I note that I am not an addressee nor
am I copied into a number of the opening emails, which are external
communications between UKGI and POL. I have no independent recollection of
this specific email chain or of the details of how the UKGI/BEIS Protocol came to
be drafted, but I have a general recollection of the protocol itself and have
reviewed my email records in order to answer Q71.
439. To situate POL00041770 in the context of what I knew and had advised upon at
the relevant time, the background to the UKGI/BEIS Protocol was that, prior to
2018, there were some concerns on POL's side that sharing certain information
with UKGI and/or BEIS could lead to the loss of legal professional privilege, or
otherwise to the uncontrolled dissemination of sensitive information. As to my
direct involvement in advising on such matters (which was limited):
439.1.0n 8 January 2017, Mark Underwood of POL emailed me asking me
whether he could share with UKGI a few short paragraphs which I had
drafted in December 2016 to go to POL’s Board by way of update ahead
of the GLO Hearing, which was listed for 27 January 2017. I advised
against that course of action as “that email contain[ed] privileged
information and sharing it [might] well waive that privilege”. I copied in
Rodric Williams in case he had a different opinion to my own. Rodric
Williams’ response indicated that he fully shared my concerns:97°
“To protect privilege we must not share any comms from our
lawyers (internal or external), or any other documents prepared
373 WBONO001068.
Page 248 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
in connection with this litigation, beyond those in Post Office with
a “need to know”. This includes our shareholder or its
representatives in government, and applies even where the
comms/documents appear benign.”
However, Rodric Williams said he was happy to set up a call with UKGI
covering any specific enquiries they might have. I was not part of that call
(if it happened, which I cannot speak to).
439.2.Later, on 17 July 2017, Melanie Corfield (POL Communications Team)
sent me two draft speaking notes for Tom Moran (Chair of the Steering
Group) for calls he was due to have with NFSP and BEIS the following
day. I advised the deletion of certain part of those speaking notes, in
particular relating to POL’s future expectations or intentions, because it
was “unclear whether th[ojse meetings would be covered by legal
privilege”.374
440. It appears from POL00041770 that on 23 February 2018 Patrick Bourke of POL
emailed Elizabeth O’Neill (UKGI), referring to a prior meeting between them in
which they had discussed the need for an appropriately structured flow of
information to UKGI in respect of the group litigation. This is the first email in the
chain. I did not attend that meeting, nor was I copied into Patrick Bourke’s email.
I can see that in response on 1 March 2018, Elizabeth O’Neill attached a
standard form protocol that she says she would be happy to amend and make
bespoke (again, I am not in copy).
441. The next email in the chain is by Rodric Williams on 27 March 2018, suggesting
that the information flow should primarily be conducted through (i) UKGI’s
374 WBONO001179.
Page 249 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
representative on the POL Board and Board Subcommittee, and (ii) regular
meetings, with provision of other documents to be considered on a case-by-case
basis, and attaching a revised protocol reflecting this.
442. In the interim, i.e. between 1 and 27 March 2018, my advice had been sought in
relation to this topic as follows:
442.1.0n 21 March 2018, Jane MacLeod emailed me and Rodric Williams
seeking our views on standard form protocol circulated by Elizabeth
O'Neill on 1 March.3”° Rodric Williams initially picked up looking at that
document.3’6
442.2.1 followed up with Rodric Williams on 22 March 2018 asking if he needed
any input from WBD on it. I reiterated that I was primarily concerned to
avoid privilege being waived in POL’s documents.°””
442.3.lt appears from my email records that some at least two calls took place
on 22 and 23 March 2018 between Amy Prime and Rodric Williams on this
subject; it is possible that I attended the latter, but I cannot be sure.°”8
442.4.0n 24 March 2018, Rodric Williams shared a revised proposed protocol
with me, Amy, Jane MacLeod and others at POL.°’° He asked Amy and
me to consider in particular the background section and Appendix B
(‘Obligations in relation to confidential information’), to “see if [this went]
far enough for privilege purposes in establishing a common interest in both
the litigation and the need to maintain confidentiality in the material we
375 POL00041684.
376 WBON0000524.
377 POL00041687.
378 WBON0000525 and WBON0000528.
378 POL00041695; POL00254174.
Page 250 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
[might] be sharing...”. My response was broadly that it “[lJook[ed] good”,
though I advised that the background section ought to capture other issues
covering the same subject-matter as the group litigation.>®°
443. As above, on the POL00041770 email chain, Rodric Williams circulated his
revised proposed protocol to UKGI, observing that information about the litigation
(including legal advice) would principally be received by Tom Cooper on UKGI’s
and BEIS’ behalf, with POL to provide UKGI’s and BEIS’ lawyers with regular
updates in meetings, and provision of documents otherwise to be considered on
a case-by-case basis.
444. Helen Lambert (UKGI) responded with a counterproposal, expressing the view
that it would be disproportionate and inexpedient to assess whether to share
information other than to Tom Cooper and via legal meetings on a case by case
basis. Although I was still not in copy at this point, Jane MacLeod forwarded me
Helen Lambert's emails on 20 April 2018.98"
445. On 21 April 2018, I responded to Jane MacLeod with some comments on the
UKGI Protocol in the following terms:
“The amended protocol gives UKGI unfettered access to information
about the Group Litigation. From a privilege perspective, that could still
be workable — privilege should still apply to the information. The risk is a
practical one — the more information that is allowed to flow to UKGI, the
greater the risk of an accidental release of privileged material.
UKGI have also significantly expanded POL's reporting requirements, in
terms of frequency and level of detail. They also want much of this in
writing rather than verbally. This could be done but it would very be
burdensome. This litigation changes shape frequently and this could
lead to weekly / fortnightly written reports. If we could water down one
380 POL00041697.
381 WBON0001236.
Page 251 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
aspect, I would go for the need to report in writing — that takes up the
most time and presents the greatest risk of a leak.”*®2
446. Thereafter Jane MacLeod, Rodric Williams and I worked on a further revised
Protocol.°®5 The three principal changes I recommended were: (i) the principle
that provision of privileged information should be dealt with on a case by case
basis should be reintroduced; (ii) the Secretary of State should be a party to
Appendix B, so that he would be bound by the requirements of that appendix in
respect of any confidential information received under the Protocol; and (iii)
adding a provision to clarify that privileged information would not be disclosed in
response to any FOI request.**4
447. The revised Protocol was sent to UKGI on 2 May 2018 (POL00041770).
Elizabeth O’Neill of UKGI responded on 11 May 2018, stating that in her view the
parties were “still a long way apart’ (POL00041770).
448. POL00041770 shows that at this point, the chain was forwarded to me and a call
was set up between me, Rodric Williams and Patrick Bourke. With reference to
Q71.2 of the Request, I have no recollection of that call, however I have an email
from Patrick Bourke to Jane MacLeod summarising what we had discussed and
agreed. I have no reason to doubt that his summary is a fair reflection of what
passed:
“We're all in the same place - ours is not an objection on principle, but is
borne of understandable concerns about how information is/would be
handled by UKGI//BEIS.
On that basis, the 4 of us discussed what it would take to give us greater
confidence, and some obvious suggestions include named people at
382 POL00041760.
383 WWBON0001240; WBON0001241.
384 WBONO001241.
Page 252 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
UKGI/BEIS, restricted channels etc. Rod, I think with Andy's help, is
working something up as a starting point ...”385
449. Thereafter, on 14 May 2018, I sent Rodric Williams some proposed wording to
feed into the draft protocol on which he was working: “On reasonable request
from the Secretary of State/UKGI, POL shall provide, within a reasonable period
of time, the information necessary to allow them to comply with their statutory or
legal duties.”386
450. On 17 May 2018, Rodric Williams sent me his latest draft, which sought to reflect
a call he had had with UKGI and incorporated a requirement for UKGI and the
Secretary of State to maintain records of recipients of confidential information (as
discussed at the call of 11 May, according to Patrick Bourke’s note of that call).
Rodric Williams asked for my comments;**” I have not identified any email which
shows I commented, so I surmise that I either did not do so or I telephoned him.
451. After this, my firm’s records do not reveal any further advice by me on the draft
protocol, which appears to have been finalised between POL and UKGI without
any further input from me.°88
M. THE GROUP LITIGATION - EARLY WORK
Part Il — EARLY INVESTIGATIONS; PREPARATION OF THE LETTER OF RESPONSE AND
GENERIC DEFENCE (Q73 to Q75, Q89)
385 POL00041772.
386 WWBONO001244.
387 WBON0001245.
388 WBONO0001251; WBONO0001648.
Page 253 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
452. This section addresses the work which my team and I undertook in order to
prepare:
452.1.POL’s LOR pursuant to the Practice Direction on Pre-Action Conduct,
served on 28 July 2016; POL00110507 (Q74).
452.2.The Generic Defence, served a year later on 18 July 2017 (Q75). However,
this section does not address the Inquiry’s Q78 concerning POL’s
pleadings in relation to the KELs database (which is instead addressed
below in Section N).
453. In order to set the LOR and Generic Defence in their proper context it is
necessary to describe some of the investigations which were undertaken
beforehand. As such, the subsections below on the LOR and Generic Defence
also include my response to Q73, as well as information on the investigative work
undertaken be Deloitte prior to service of the Defence. Lastly this section
addresses the Inquiry’s Q89, concerning the report Deloitte produced in
September 2017, shortly after service of the Generic Defence.
(i) The Letter of Response (Q73 to Q74)
Overall approach — division of labour and sign-off process
454. The LOR (POL00110507) was the product of around three months of dedicated
work on the part of myself and my team, with input from Counsel, POL and
relevant external parties (such as Deloitte and Fujitsu). For that reason, it is
impractical to attempt to exhaustively set out my involvement in its drafting and
any advice I gave to POL as to its response. I summarise what I believe are the
key points below.
Page 254 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
455. The drafting of the LOR was broken up into sections that were assigned to
particular members of my team at WBD to investigate and/or draft;38° save where
Counsel drafted sections of the LOR, it was WBD (led by myself) that held the
pen. I was assisted by Tom Porter, Amy Prime, Paul Loraine, Andrew Pheasant,
and Jonny Gribben.
456. In essence, my role was to coordinate the whole process of drafting the LOR. In
particular:
456.1.l managed the programme of work involved in preparing the LOR to ensure
that drafting and review deadlines were met by the WBD team and
others.5 By way of example, a first draft of the LOR was produced and
sent to Rodric Williams for his review on 27 June 2016.99"
456.2.Where my team had queries or required advice as to strategy in relation
to their assigned sections, I was the main point of contact. Where
appropriate I made decisions on how to respond to parts of the LOC.%% I
oversaw legal research conducted by my team and provided comments
on their research notes.°%° Where I came across any information that I felt
was relevant to the drafting of the LOR I fed that to the relevant members
of my team.3%
456.3.1 took instructions from the POL legal team throughout, and, for decisions
that needed to be made by the Steering Group, WBD prepared papers
389 POL00242335.
38 See for example: POL00242335 (workplan as at 6 June 2016); POL00243124 (workplan as at
8 July 2016).
391 WBONO001019.
382 WBON0000431.
383 WBON0000427.
394 For example, in relation to the Suspense Account bug, see WBON0000424.
Page 255 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
which outlined the decisions that had to be made and our
recommendations.°® I oversaw these papers which reflected my
recommendations for the best course of action (as well as seeking the
input of the Counsel Team),°° particularly for handling trickier points in the
LOR from a commercial or tactical standpoint.
456.4.1 liaised with Counsel where their input was required to draft or review
sections of the LOR. Throughout the process I worked closely with our
Counsel Team, Tony Robinson QC and Owain Draper, to discuss the
developing draft and our factual and legal analysis.°°” Counsel held the
pen on specific sections of the LOR which involved more complex legal
points, namely Section 4 (legal duties) and Section 6 (heads of claim).°°°
I also liaised with Brian Altman QC in relation to his review of historic
charging decisions by POL as this was relevant to the misfeasance and
malicious prosecution claims intimated in the LOC and he looked over the
relevant sections of the LOR in draft.%°° I deal further with Brian Altman
QC’s review below at §§458-465 in the context of Q73.
456.5.My team and I also obtained input from Fujitsu and Deloitte on parts of
Section 5 (concerning Horizon defects, and data integrity and remote
access), Deloitte having by now been instructed to prepare an interim
report on the remote access point in advance of the deadline for the
LOR.“ I deal with (i) Deloitte's preliminary investigation, and (ii) the
395 See for example: POL00243114.
39 WBON0001025; = POL00025373.
397 See for example WBON0000426; WBONO0001021; WBON0001024; WBONO0000432;
WBON0001031; and WBON0000434.
398 WWBONO0001025.
399 WBON0001023; WBON0001033; WBON0001047.
400 WWBON0000423, POL00041238.
Page 256 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
drafting of the data integrity and remote access section further below at
§§466-472 and §§473-478, respectively. Similarly, we consulted
extensively with individuals within POL and other organisations who had
previously been involved (for example, CMS Cameron McKenna) in order
to gather relevant factual information. That input was largely obtained by
the individuals within my team who were responsible for the relevant
sections.*°"
457. The process for POL to review and approve the LOR was carefully coordinated.
At the Steering Group meeting on 20 July 2016 (where the first draft of the LOR
that was ready for the Steering Group’s review was discussed), it was agreed
that: “Where a statement of fact is made in the LoR, BD are to ascertain its
provenance”.*°2 Input was therefore sought from key subject matter experts at
POL, together with the in-house legal team and the members of the Steering
Group (who themselves had extensive knowledge of the underlying subject
matter relevant to their parts of the business). The process of review and sign-
off by POL proceeded, in short, as follows:
457.1.An initial draft of the LOR was produced and sent to Rodric Williams for
review on 27 June 2016.4 Thereafter the draft underwent further revision
by the WBD and counsel teams, and the first working draft that was ready
for the Steering Group’s review was circulated to POL on 16 July 2016.4
That draft was substantially complete but for two sections, namely (i) data
integrity and remote access (as enquiries with Deloitte and Fujitsu were
401 See, by way of example, WBON0001028.
402 POL00243355.
403 WWBONO001019.
404 WBON0000435.
Page 257 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
ongoing at this point, as explained further below); and (ii) the GLO, as this
remained the subject of ongoing correspondence with Freeths.
457.2.Separately and in parallel, on 18 July 2016 I circulated the draft LOR to
Kathryn Alexander and Shirley Hailstones, who had led POL’s in-house
investigation team under the management of Angela Van Den Bogerd
during the Mediation Scheme (see §140.5 above), and had deep
experience of working in branches and operating Horizon. Their
involvement was sought in response to the perceived need to validate
even basic facts relied upon in the LOR, as is reflected in my covering
email to them: “[pjlease err on the side of caution — if you're not sure if
something is correct, please flag it."4°
457.3.On 19 July 2016, I sent the draft LOR to Jessica Madron (Head of Legal
at POL), drawing her attention to particular sections of the LOR concerning
contractual issues, in particular the terms we accepted were implied into
the SPM-POL relationship .4°%
457.4.Each member of the Steering Group was to review the sections marked
for their attention in a spreadsheet I attached to my covering email (which
allocations correlated with the sections of the business each member was
responsible for).4°”
457.5.The above individuals responded with their comments which I collated with
the assistance of the WBD team.
405 WWBONO0001038.
406 WWBONO0001036; see further our follow-up exchanges at WBON0001048 and WBON0001054.
407 WBON0000436.
Page 258 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
457.6.The GLO section, as approved by Tony Robinson QC, was circulated to
the Steering group for comment on 25 July 2016,4°° and subsequently fed
into the draft. I discuss the process by which the data integrity and remote
access section was drafted below, but in short it was finalised on 27 July
2016 and approved by Tony Robinson QC on 28 July 2016.
457.7.The complete, finalised draft of the LOR was circulated to the Steering
Group for on 27 July 2016 (POL00041259), and Jane MacLeod gave
instructions on behalf of POL to send it the following day.*°°
Investigations undertaken — Brian Altman QC’s Review
458. By July 2016 POL had either completed Tim Parker's implementation of the Swift
Review recommendations or migrated them to be done within the group litigation
preparation work (see above, §§417-423). Brian Altman QC’s review (which was
commissioned to implement Recommendations (1) and (2) of the Swift Review)
was ongoing at that time, and his work was therefore carried on and completed
in the context of the group litigation to help POL to understand the risks around
any possible malicious prosecution claims that the Claimants may bring.
459. The form of his review was to examine the prosecutions of a number of individual
SPMs who had been charged with theft and false accounting simultaneously, in
order to identify (i) whether there had been insufficient evidence to charge theft
such that the relevant individuals had been improperly pressured to plead guilty
to false accounting, and (ii) whether there was a pattern of behaviour on POL's
part in this respect.
408 WBONO001050.
409 WBONO001061
Page 259 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
460. In order to inform our response to parts of the Claimants’ LOC alleging malicious
prosecution and misfeasance in public office, Brian Altman QC was asked to
complete his review if possible in advance of the LOR and failing that, to cast his
eye over the relevant sections of the draft LOR (which he did).4"°
461. My email records indicate that Brian Altman QC sent his advice initially on 25
July 2016.41 I have no specific memory of this, but it is evident that I reviewed it
at the time, as I responded the following day with one query on the substance of
his advice that overlapped with one of the civil claims advanced in the group
litigation, and a request regarding its form (namely, that he reflect within it that he
had originally been commissioned by Tim Parker but was now instructed to
continue his review for the purpose of assisting POL’s defence in the group
litigation). My query on the substance was whether POL could properly maintain
that it was not bound by the Code for Crown Prosecutors in the LOR, without
opening itself up to an argument in the criminal sphere that proceedings brought
by it were abusive.*'? Brian Altman QC responded in the affirmative, provided
that “there was no wholesale policy to disapply the Code as and when it suited
POL's own ends” (which in his view, there was not).4*? He provided me with a
finalised version of his advice an hour or so later.4"4
462. Upon receiving the finalised version on 26 July 2016, I circulated it to Rodric
Williams, Jane MacLeod, Patrick Bourke and Mark Underwood at POL,
summarising his conclusions (POL00022754; POL00112884).*' By Q73, I am
410 See above, §456.4.
411 POL00408673.
412 WBON0000444.
413 POL00408673..
414 WBON0000445; WBON0000446.
415 I do not understand the reference in Q73.1 to the advice being redacted. I do not recall there
being a redacted version of this advice and according to my records, the version that I reviewed
Page 260 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
asked to comment on those documents. I do not now specifically recall reviewing
the advice, though I broadly remember Brian Altman QC doing this work,
preparing the report, and concluding that he had not seen any evidence of
malicious prosecutions. I recall that my general view was that the conclusions
emerging from his review — namely that the primary and secondary allegations
against POL (as recorded in POL00022754) were misplaced — aligned with and
supported the approach we had taken when drafting the LOR. Beyond this my
thoughts on Brian Altman QC’s advice at the time would have been as set out in
POL00022754.
463. Specifically, I am asked to comment on a quote taken from paragraph 208 of the
advice which reads: “but that POL has been using the criminal justice system as
a means of enforcing repayment from offenders by charging and pursuing
offences that will result in confiscation and compensation orders”. I recall that
occasionally people within POL would talk about using private prosecutions as a
means to recover debts owed, but POL’s legal team (in particular, Rodric
Williams) repeatedly pushed back on this, and were adamant that prosecutions
could not be used in this way and that each one needed to be brought on its own
merits.
464. Reading paragraphs 208 to 211 of the advice, in my view the above quotation in
the Request has been truncated and the effect of that is to take it out of context.
My reading of paragraph 208 is that Brian Altman QC was laying out the
background that other persons had criticised POL for using the criminal process
and circulated was unredacted. Further, with reference to Q73.4, I do not recall briefing anyone
else on the advice beyond the email in POL00022754. My firm’s records show that I sent the
advice to Paul Loraine, an associate in my team who had assisted with drafting Brian Altman
QC’s instructions (WBON0000470); and that I also asked Amy Prime to add it to our internal
file (WBONO0000450).
Page 261 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
to recover monies owed to it. I note that his footnote 25 cites the Claimants’ LOC
as a source for this. I did not read this paragraph as saying that these were Brian
Altman QC’s thoughts, and that he was criticising POL for this. My reading of
paragraph 208 is then consistent with the following paragraphs where Brian
Altman QC observed that (i) it is appropriate to consider the orders that might
follow conviction (including confiscation orders) when indicting offences, and (ii)
in each of the 8 cases he reviewed, there was a proper legal and evidential basis
for seeking confiscation orders. In my opinion, my email accurately reflected the
advice on this point as I understood it (i.e. in the paragraph “The secondary
allegation (that offences... follow conviction)”).
465. In my assessment, therefore, the part of Brian Altman QC’s advice to which I am
referred (and indeed the advice as a whole) did not mandate further action from
the perspective of defending threatened malicious prosecution claims in the
group litigation. That this was my view at the time is reflected in the final
paragraph of my email. The approach we were taking in the LOR was already
consistent with Brian Altman QC’s findings in his review, so no changes to
relevant sections of the LOR were needed.
Investigations undertaken — remote access
466. Following Tony Robinson QC’s advice in conference on 9 June 2016, WBD
instructed Deloitte on a privileged basis to investigate the remote access point —
in the short term, with a view to assisting POL in formulating its response to the
Claimants’ LOC. To this end, Deloitte was to produce a preliminary report by 28
Page 262 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
June 2016, with further testing to be conducted thereafter to provide reassurance
and inform the conduct of POL’s defence to the litigation thereafter.41°
467. The background in terms of what was then understood to be the case (at least
by me) in relation to remote access is set out above at §§202 ff taken together
with §§299-302.
468. Deloitte provided a draft of their preliminary report on 8 July 2016.41” Deloitte
prepared this report having met with people at POL and Fujitsu, reviewed
documentation about Horizon and undertaken various testing on the system (as
more particularly detailed in the report itself). WBD did not set the scope of those
enquiries or oversee them; for example, WBD did not attend the meetings
between Deloitte and Fujitsu. In my view, these were enquiries into the deep
technical details of Horizon, and Deloitte had been engaged because they had
the technical expertise to critically challenge the information being provided by
POL and Fujitsu.
469. In relation to Balancing Transactions, the report identified (inter alia) that:
469.1.Balancing Transactions are “exceptional processes used by Fujitsu
support staff to correct exceptional errors in system processing/fix issues
or bugs in the recording of data” the use of which “is very rare”.
469.2.Writes by Fujitsu support staff to the BRDB to implement a Balancing
Transaction flow to the Audit Store and such staff “cannot amend the
related audit files’. They can only insert Balancing Transactions and “will
not have any privileges to update or delete records in the database”.
416 WBONO001016.
417 WBON0001644; POL00243100.
Page 263 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
469.3.The known lifecycle of the system is “predominantly limited to HNG-X due
to previous Audit Store retention limitations”.
469.4.In Horizon Online “there have been several hundred instances of
Balancing Transactions’, but only one (the previously identified use of a
Balancing Transaction in 2010) was to address a discrepancy caused by
a bug, with the others all being to unlock stock units.
470. In relation to privileged user access, the draft preliminary report confirmed that:
“[aJt various layers of the Horizon infrastructure there exist accounts with
privileged access rights which could be used to modify or insert data relevant to
transactions at branches should they not be adequately controlled”. This could
include the unauthorised use of the Balancing Transaction process, for example
by “a superuser account’ on the Oracle DB (this being “the nucleus of the
[BRDBJ’ and Balancing Transactions inserted in the way outlined above
effectively being “a specialised ‘legitimised’ way of using such Oracle access” by
Fujitsu Support). Further work was required to identify and assess the
capabilities of privileged users to create, edit or delete branch accounting data in
this and other ways (for example, within the Audit Store itself as identified
previously).4"8
471. Whilst Deloitte’s account of the conventional use of the Balancing Transaction
process was similar to what we had previously understood (albeit that it was not
clear whether this or a similar functionality had existed in Legacy Horizon), its
report crystallised that privileged users at Fujitsu with access to the BRDB could
delete and modify existing transaction data. This was problematic for POL as it
418 POL00243100.
Page 264 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
meant that its “historic statements about not being able to edit or delete
transactions appear[ed], at least on face value, to have been materially incorrect”
and this placed POL's limitation defences at risk and created a basis for
unwinding previously settled cases.*1°
472. I conveyed these findings to POL, noting that privileged access to the BRDB “is
subject to strict controls and Deloitte’s current understanding is that it would not
be possible to delete or edit transactions without leaving a footprint in the audit
trail’, as well as Deloitte’s view that “(i) this type of access is not unusual and (ii)
the likelihood of someone actually making such changes is extremely low’. I
advised that it was not yet clear what impact, if any, there could be or had been
on branch accounts (including those of the Claimants) as a result of these types
of access rights, and that further investigative work by Deloitte would be
required.42°
Drafting and sign-off of the remote access section
473. By Q74.3 of the Request, I am asked to explain the basis on which paragraph
5.16 of the LOR (POL00110507) was drafted and approved.
474. Paragraph 5.16 sets out the four ways in which (it was then thought) POL or
Fujitsu could influence branch accounting data. In the LOR as sent, it reads:
“5.16 Transactions which make up the branch accounts are generally
generated in branch. There are however four ways in which Post Office
(or Fujitsu on Post Office's instruction) can influence those accounts:
419 WBON0000430. In addition to cases being settled through the mediation scheme, POL's
Network Transformation programme required SPMs moving to the new contract structure to
release historic claims against POL. It was therefore expected that many Claimants in the
litigation would be subject to settlement agreements and would be looking for a way to unwind
those agreements.
420 WBON0001030.
Page 265 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
5.16.1 Transactions originating at Post Office. A number of
"transactions" are generated by Post Office and sent to
branches, namely transaction corrections, — transaction
acknowledgements and remittances of cash / stock into a
branch.®*3 A key feature of these transactions is that they must be
approved in branch (by the postmaster or his assistants) before
they form part of the branch accounts.
5.16.2 Global Users. Global Users are setup by default on Horizon in
every branch. These are user accounts for Post Office staff to
use when undertaking activity in a branch, such as training or
audits. It is possible for these Global Users to conduct
transactions within a branch's accounts. However, this access
is only possible if the user is physically in the branch using a local
terminal and the transactions are recorded against the Global
User ID.°4
5.16.3 Balancing transactions. Fujitsu (not Post Office) has the
capability to inject a new "transaction" into a branch's accounts.
This is called a balancing transaction.°5 The balancing
transaction was principally designed to allow errors caused by a
technical issue in Horizon to be corrected: an accounting or
operational error would typically be corrected by way of a
transaction correction. A balancing transaction can add a
transaction to the branch's accounts but it cannot edit or delete
other data in those accounts. Balancing transactions only exist
within Horizon Online (not the old version of Horizon) and so
have only been in use since around 2010.% Their use is logged
within the system and is extremely rare. As far as Post Office is
currently aware a balancing transaction has only been used
once*’ to correct a single branch's accounts (not being a branch
operated by one of the Claimants).°°
5.16.4 Administrator access to databases. Database and server
access and edit permission is provided, within strict controls
(including logging user access), to a small, controlled number of
specialist Fujitsu (not Post Office) administrators. As far as we
are currently aware, privileged administrator access has not
been used to alter branch transaction data. We are seeking
further assurance from Fujitsu on this point.
[53] See paragraph 7.16 onward in Second Sight's Part One Report for
a more detailed explanation of these processes.
Page 266 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
[54] Strictly speaking, the Global User ID should be used to generate a
new unique ID for the Post Office staff member and the new ID would
then be used for training, audits, etc.
[55] The use of balancing transactions was explained to Second Sight
and is referenced in its Part Two Report at paragraph 14.16.
[56] Post Office is making enquiries as to whether something akin to a
balancing transaction existed in Horizon before the upgrade in 2010.
[57] This was in relation to one of the branches affected by the
"Payments Mismatch" error described in Schedule 6.
[58] Several hundred other balancing transactions have been used but
not in a manner that would affect branch accounting. These were
generally used to "unlock" a Stock Unit within a branch.”
475. This text was approved by me and Tony Robinson QC, and was signed off by the
Steering Group, POL senior executives, and Jane MacLeod. As I explain further
below, Tony Robinson QC and I would have preferred a clearer and more direct
statement that administrator or ‘privileged user’ access could potentially be used
to change branch accounting data (and the Steering Group was aware of this).
Nevertheless, we were satisfied that the above text (together with the footnotes)
reflected the factual position as we understood it to be following receipt of
Deloitte’s preliminary report whilst allowing for the fact that there were a number
of respects in which the picture was not settled and further investigations by
Deloitte were required (for example, as to whether there was an equivalent of the
Balancing Transaction tool in Legacy Horizon and if so, whether this had ever
been used; and likewise, whether administrator access to the BRDB had ever
been used in practice to influence branch accounts).
476. In order to give the Inquiry a full picture of how paragraph 5.16 came to be drafted
(including who contributed to the wording and what considerations were in play
during the review and sign-off process), I have reviewed relevant documents in
my firm's file and can comment as follows:
Page 267 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
476.1.Based on my firm’s records, after receipt of Deloitte’s interim report I
worked up a preliminary draft which was structured around the various
forms of remote access SPMs had alleged during the Mediation Scheme,
identifying what was and was not possible by reference to those
allegations." I sent this to Tony Robinson QC on 20 July 2016422 and he
gave a clear steer that we needed to plainly and separately identify the
different types of remote access that we now understood to be possible
(i.e. as opposed to responding by reference to the various allegations that
SPMs had previously made about remote access).42°
476.2.1 accepted Tony Robinson QC 's advice and produced a restructured and
more detailed draft on 21 July 2016.44 This ended up forming the basis of
what ultimately became paragraph 5.16 in the LOR; indeed, regarding
‘transactions originating at Post Office’, ‘Global Users’, and ‘Balancing
Transactions’, there is little difference between my 21 July 2016 draft and
the eventual LOR.*?5 I focus therefore on the description of administrator
access. As to this, my draft of 21 July 2016 read:
“Access to databases. There are a small number of persons at
Fujitsu (not Post Office) who have special permissions to access
and edit, within strict controls, the core databases that sit behind
Horizon. Use of these permissions is logged and so it is believed
that there would be an audit trail of any activity undertaken using
these permissions. Enquiries are continuing as to whether this
421 WBONO001041.
422 WBONO0001040.
423 WWBON0000438.
424 WBON0001042; WBON0001044.
425 The main differences being that, in relation to ‘transactions originating at Post Office’, a
sentence to the effect that Transaction Corrections etc are an everyday occurrence was
removed; and in relation to ‘Balancing Transactions’, the underlined words were added to “[t/heir
use is logged within the system and is extremely rare’, and a footnote was added to reflect that
enquiries were being made to confirm whether or not a similar functionality existed within
Legacy Horizon. These minor edits were all suggested by Tony Robinson QC on 21 July 2016:
see his draft at POL00029997, and his covering email at POL00408665.
Page 268 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
access could be used to affect a branch's accounts but we
currently understand that, even if this is possible, it would be a
difficult and time consuming process. Moreover, given the above
methods open to Post Office to deal with errors in a branch's
accounts, the use of this access to amend a branch's accounts
would be extremely rare — indeed, Post Office is making enquiries
as to whether it has ever happened.”*26
476.3. Tony Robinson QC agreed this wording with only minor linguistic tweaks*?”
(which, based on my emails, we discussed by telephone on 21 July 2016
although I do not remember the call).478
476.4.1 circulated our agreed draft to the Steering Group the same day in
advance of a call that evening to discuss the draft LOR.4”° In my covering
email, I asked the Steering Group to bear the following in mind when
reviewing the wording:
“1. In light of comments yesterday, we've provided a slightly
longer explanation so to hopefully present this issue in a better
light.
2. Tony agrees with the current wording but has reiterated the
importance of dealing with this point candidly, even if that
does cause some short-term pain.
3. We do not yet have a 100% clear picture on some of the
technical and operation issues on this topic. We therefore
need to be careful not to overstate our case. This draft wording
will also need to be run past Deloitte / FJ” (emphasis added).*9°
426 WWBONO0001044.
427 Compare my 21 July 2016 draft (WBON0001044) with our agreed draft of the same date which
I subsequently circulated to the Steering Group (POL00243366). Prior to our call on 21 July,
Tony Robinson QC suggested that we could specify that POL had only recently become aware
of privileged user access (cf. his draft at POL00029997). It is clear however that we ultimately
decided against that addition.
428 WWBON0000439.
429 POL00243366.
430 POL00024801.
Page 269 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
476.5.When the Steering Group met on the evening of 21 July, I recall there
being a lot of discussion about how POL should frame the ‘privileged user’
access point above. I remember there being disquiet on POL’s part at the
prospect of contradicting their previous public statements on this point
(POL having previously denied it was possible to edit or delete transaction
data).431
476.6.Following the call, Rodric Williams circulated the wording that the Steering
Group landed on:
“Database and server access and edit permission can be
provided, within strict controls, to a small, controlled number of
specialist Fujitsu personnel. Use of these permissions is logged
but rare. Enquiries are continuing as to whether this [particular
form of] access could be used to affect a branch's accounts, and
if so, whether this has happened" 452
476.7.1 tweaked this by replacing “can be provided” with “is provided” (emphasis
added) in the first sentence, as by this point, Deloitte had confirmed the
existence of privileged users at Fujitsu with this type of access. This
reflected the strong advice I had received from Tony Robinson QC (and
communicated to the Steering Group) that it was important to be as
transparent as possible.*°%
476.8.1 circulated an updated version of the data integrity and remote access
section incorporating this revised wording to Fujitsu and Deloitte on the
evening of 21 July 2016.4% Given the approaching deadline for the
431 As is also reflected in the Steering Group’s email discussion preceding the meeting:
WBON0000441. This chain also contains various alternative formulations suggested by
different members of the group in the run-up to the meeting.
432 POL00024876; words in square brackets added after the meeting by Mark Davies (POL's
Communications Director), see: WBON0000440.
433 POL00024876.
434 WBON0001045 (Fujitsu) and POL00408671 (Deloitte).
Page 270 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
submission of the LOR, I asked both to revert with any comments by close
of business the following day.
476.9.Mark Westbrook of Deloitte responded that the draft it “/sJeems to reflect
our understanding, although I’m not sure we got to the bottom of whether
balancing transactions were available pre-Horizon online”.4*> By contrast,
Fujitsu did not respond until 27 July 2016, the day before the LOR was
due.4® When they did so, they expressed the “interim view” that:
“we don’t believe the key element regarding the sanctity of the
Core Audit Process comes through in your proposed response
and this remains the cornerstone of our inputs to date and the
presentations we’ve done on this point in the past to Post Office
and Bond Dickinson. ... the Core Audit Process captures every
submitted “basket” accurately and without error and_cannot
subsequently be changed but only added to (all such additions
include details as to what created the addition etc. to create a full
audit log).°"
476.10. This was a frustrating response to receive, given it was raised shortly
before the LOR was due to be sent and it was inconsistent with Deloitte’s
findings that privileged users at Fujitsu could delete and modify transaction
data. I communicated to Jane MacLeod that I did not think that the
“disconnect in understanding’ between Fujitsu and Deloitte could be
resolved before the deadline for the LOR had passed, and that (erring on
the side of caution) we ought to prefer Deloitte’s view and seek Fujitsu’s
sign-off only on other parts of the LOR.4%8
435 POL00243580.
436 WBON0000442; WBON0000447.
487 WBON0000449.
438 POL00023428.
Page 271 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
476.11. Meanwhile, those at POL continued to deliberate the proposed wording.
There was continued unease (including among the General Executive)
about contradicting POL’s previously stated position, and about
acknowledging that “enquiries [were] continuing as to whether [privileged
user] access could be used to affect a branch's accounts, and if so,
whether this has happened’.**° My advice was that POL would not have a
complete picture of the situation before the deadline for the LOR but that I
“[had] in mind Tony's strong advice about being transparent on this point
as far as possible” .44°
476.12. A call was arranged for 27 July 2016 for the Steering Group to give me
clear instructions on the wording for the LOR.‘*' The finalised wording
approved by the Steering Group was as follows (tracked against the earlier
version circulated to Deloitte and Fujitsu):
“Administrator Aaccess to databases. Database and server
access and edit permission is provided, within strict controls
(including logging user access), to a small, controlled number of
specialist Fujitsu (not Post Office) perseaneladministrators. Use
£ thy rn. ns-islogged-butrare-Enauirias are ptinuin.
PY gg AG
as_we are currently aware, privileged administrator access has
not been used to alter branch transaction data. We are seeking
further assurance from Fujitsu on this point.”42
488 See for example: POL00025320; WBON0000448; POL00024824.
440 POL00024794.
441 POL00024828.
442 POLO0357378. See emails from Steering Group members signing off: WBON0000452 (and my
response WBONO0000453); WBONO0000454; WBON0000455; WBONO000456;
WBONO0001057. WBON0000456
Page 272 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
476.13. This revised formulation was then circulated to both Fujitsu and Deloitte
after the Steering Group call.443 Mark Westbrook responded that he was
“absolutely fine” with these amendments.*“4 Fujitsu responded later that
day on the LOR but did not respond substantively on the draft wording
around remote access.4>
476.14. I forwarded the finalised wording to Tony Robinson QC, indicating that!
had advised the Steering Group on the importance of transparency on the
privileged user access point, and that I had indicated a strong preference
for “a more direct statement that the permissions could potentially be used
to change branch accounts” (as our previously agreed wording had been).
Nevertheless, I felt that we could “live with” the client’s preferred wording
as it was sufficiently accurate and make clear that enquiries were
continuing.446
477. \n terms of final sign-off on the LOR, including paragraph 5.16 as approved by
the Steering Group, I circulated a final draft on the evening of 27 July 2016
(POL00041259).
478. POL00022663 is an email sent by Jane MacLeod to Paula Vennells and Alisdair
Cameron on 28 July 2016 which I do not believe I have seen before (and similarly
443 POLO0408686; WBON0001672 (Fujitsu) and WBON0000457 (Deloitte).
444 WBON0000458. He added: added, “/ don’t know if you could strengthen your position further
in relation to 1.3.4 with wording to the effect of ‘such database access being a necessary
requirement of IT administration and support any IT system’. Or similar—to (correctly) normalise
it.” I responded that “We toyed with that extra sentence but then thought that if it was very
“normal” the question might be... Well why didn’t POL/FJ know about it sooner? So we kept our
powder dry on that one” (POL00408699). This reflected my earlier advice to POL that all historic
statements on remote access should be investigated, but that there was insufficient time to do
that before the LOR was submitted: POL00024794.
445 WWBONO001055.
446 POLO0408688.
Page 273 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
its attachment POL00022664 is an internal POL email chain which does not
feature within my firm’s files). I can see from POL00022664 that Alisdair
Cameron and Jane MacLeod had some reservations about the second and third
sentences of the sub-paragraph on privileged user access, but that Jane
MacLeod sought to reassure Alisdair Cameron and Paula Vennells that the
proposed wording had been agreed by the Steering Group as necessary from a
“transparency” perspective, otherwise POL might be “challenged for under-
stating a potential risk’. I do not have access to the rationale she attaches to that
email, so I cannot comment on whether it coheres with my recollection of events.
In POL00022663, Jane MacLeod refers to the description of privileged user
access as having been arrived at by the Steering Group after “much discussion”
(which I would agree with), and comments that the letter is ready for service
subject to Paula Vennells’ and Alisdair Cameron’s comments. I infer that they
approved the LOR since, as noted above, I received Jane McLeod's
authorisation to serve the LOR later that day.*47
(ii) The Generic Defence (Q75)
Overall approach — division of labour and sign-off process
479. POL’s Generic Defence was served on 18 July 2017, around a year after the
LOR. The Claimants had served their GPOC on 23 March 2017, although I was
aware of its substance before then as they provided a draft GRPOC in December
2016. As with the LOR, the drafting of the Generic Defence was a lengthy and
involved process, supported by further detailed factual investigations. On 6 July
447 WBONO0001061.
Page 274 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
2017, the Claimants provided a draft Amended GPOC which was taken into
account in the Generic Defence, meaning that there was no need for POL to
make an amendment to the Defence when the Claimants’ Amended GPOC4®
was formally served.
480. As is usual in significant civil litigation, the Generic Defence was settled by the
Counsel Team (Tony Robinson QC and Owain Draper) with the WBD team
playing a supporting role. It was therefore Counsel who ultimately determined
the structure of the Generic Defence and made decisions as to the granularity of
pleading.**° I was involved in those discussions but was not drafting the
document, and I followed Counsels' steers in this respect.45°
481. In terms of the process by which the Defence was prepared and my involvement
in that process (Q75.1), as with the LOR it is only feasible to set out the headline
points in this statement. I believe these are as follows:
481.1.As noted above, in July 2016 Deloitte produced a preliminary report on
remote access to assist with our response to the LOC. Thereafter they
were instructed to continue with these investigations. Jonny Gribben led
on this work, liaising with Deloitte, Fujitsu and POL as required. These
investigations are, so far as relevant, described below at §§483-490.
481.2.0n 14 February 2017, a Steering Group paper prepared by WBD collated
a list of issues on which evidence was still required as they had “so far
garnered little attention or on which the position remain[ed] unclear’.4°"
This was further refined when the Claimants’ GPOC was formally served
448 POL00250455.
449 WBONO0001081.
450 WWBONO001071.
451 WBONO001677.
Page 275 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
on 23 March 2017. Work to investigate these matters was led by members
of my team, who liaised with POL and external parties as necessary.*5?
Kathryn Alexander and Shirley Hailstones at POL, in particular, worked
with my team to investigate matters on which POL’s input was required,
since they had deep knowledge of POL’s internal processes and the way
branches operated (see above, §140.5 and §457.2).4°3
481.3.1 largely supervised my team’s work in this regard, but led (with Jonny
Gribben) on progressing the investigation into the issue of unmatched
balances on POL’s general suspense account and the extent to which
these could be linked with SPM shortfalls (this was the subject of Jonathan
Swift QC’s Recommendation (8)). A list of questions was prepared for
Deloitte to consider and they were instructed in March 2017.4% Their initial
findings were produced on 16 May 2017;45° I understood that there was
some connection between POL's suspense accounts and branch
accounts, and so an error in the suspense accounts could possibly cause
an error to flow through to the branch accounts — although Deloitte's work
had so far not reached that conclusion. They were instructed to continue
with their work, the progress of which we kept under review.
481.4.The bulk of the drafting work was done in May and June 2017. Owain
Draper produced a draft Opinion on the burden of proof (which was of
central importance to the question of who was legally responsible for
shortfalls in branches), which formed an important starting point for the
482 WBON0000474.
453 WWBONO0000478.
454 POL00023448; POL00023449;
485 WBON0001079; WBON0001080.
Page 276 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Generic Defence. The Counsel Team sent me various outlines and drafts
during this period which I reviewed and commented on. In support of their
drafting work:
(i) The WBD team sent briefing notes to Counsel on the factual
investigations that had been undertaken to date and carried out legal
research as required.*5° Counsel were also provided with Freeths’
response to a Request for Further Information we had made on 27 April
2017 when it arrived (in May 2017).
(ii) On 7 June 2017, a meeting took place between me, Counsel, and Elisa
Lukas (WBD), and Kathryn Alexander and Huw Williams (who worked
in Kathryn Alexander's team and who, I recall, was also a former SPM)
at POL, so that Tony Robinson QC could speak with them directly about
POL’s branch operating practices.*°”
(iii) A meeting took place between me, Fujitsu, Counsel, and Elisa on 22
June 2017, for the purposes of understanding more about Horizon in
relation to pleading certain points of the Generic Defence.*® My
expectation was that Gareth Jenkins would attend for Fujitsu along with
Torstein Godeseth and Pete Newsome, but I was informed on 20 June
2017 that Pete Newsome had not invited him to attend. I asked Rodric
Williams to communicate to Fujitsu that if Gareth Jenkins did not
attend, they ought to “be sure that Torstein w[ould] be able to answer
the questions.”459
456 See for example: WBON0001085; POL00249406; WBONO001116.
487 WBON0000481.
458 WWBONO001112.
459 WBONO001115.
Page 277 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(iv)l, my team, and Counsel would regularly have calls to discuss the
developing draft and exchanged comments by email.4©° Rodric
Williams was kept abreast of developments and was sent partial drafts
to review and comment on.*61
481.5.A complete first draft was produced by Tony Robinson QC on 4 July 2017,
which he sent to me for review.*® In the week following that:
(i) I reviewed the points that Tony Robinson QC had left to be checked in
that draft, some of which I asked Amy Prime and Elisa Lukas in my
team to resolve, others I managed myself.‘ I also attempted a first
draft of the executive summary section.
(ii) The draft was provided to Rodric Williams who returned his comments
on 7 July 2017.46
(iii) Additionally, on 4 July 2017 Fujitsu*®® and Deloitte*** were sent
relevant extracts from the Defence dealing with Horizon for their
review. I joined a call with Pete Newsome and Torstein Godeseth of
Fujitsu on 7 July following which my team made some amendments to
the draft which Fujitsu were then asked to further consider.*®” Deloitte
and Fujitsu's input is (so far as relevant) considered further below at
§§483-490, §§493-494, and §499.
460 See for example: WBON0000484; WBONO001126.
461 WBON0001083; © WBONO001121.
462 WWBONO0000485.
463 WWBON0000487 and WBONO0000489.
464 WBON0001145; POL00249670.
465 WWBON0000491; WBON0000492.
465 WWBONO001128.
487 WBONO0001147.
Page 278 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(iv) Relevant parts of the draft were circulated to subject-matter experts at
POL to confirm their accuracy, including: Kathryn Alexander and
Shirley Hailstones, who, as explained above, had assisted my team
with various investigations in connection with the LOR (they were sent
the description of the parties, POL’s processes, and the NSBC
Helpline);4°8 and Gayle Peacock, POL’s Head of Branch Support (who
was sent the section concerning the NSBC Helpline).*® After they
approved these sections, Angela Van Den Bogerd was then sent them
for a second-line review.4”° Kathryn Alexander's, Shirley Hailstones’
and Angela Van Den Bogerd’s input is considered further below in the
context of paragraph 43 of the Generic Defence (§§500-509).
(v) My team fed in comments to a Master draft which I kept under review.
Those comments that were received by 9 July 2017 were reflected in
the draft. The updated version of the draft was provided to Counsel the
following morning.471
481.6.On 12 July 2017, Tony Robinson QC circulated a complete draft which was
ready for review and approval by POL.*”? This process was slightly
complicated by the fact that we were still awaiting Deloitte’s report on
POL’s general suspense account (which was relevant to our response to
paragraph 38 of the Amended GPOC), as I set out further below. The
review and sign-off process at this stage was:
468 WBON0001130; WBONO0001132.
469 WBON0001138; WBONO0001140.
470 WWBON0000493; WBON0000492.
471 WBON0000496. See also: WBONO0000497.
472 WWBONO001156; WBON0001157; WBONO0001683.
Page 279 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(i) The topic of the Generic Defence was discussed at a meeting of the
Steering Group on 12 July 2017 and later that day I circulated the draft
Generic Defence to them for comment on.4”
(ii) Outstanding comments from subject matter experts at POL (in
particular, Angela Van Den Bogerd)*”4 and Deloitte*”® were fed in.4”6
Fujitsu were asked to review the section relevant to them as updated
following the call on 7 July 2017; as set out below at §§493-494 and
§499, they confirmed their agreement on 12 July 2017 subject to minor
comments.4””
(iii) I believe the draft Generic Defence was discussed at a meeting of the
General Executive on 13 July 2017 (which I believe I attended but I do
not recall the meeting).478
(iv)On 13 July 2017, Mark Underwood informed me that Deloitte’s
suspense account report was not going to be ready before the
deadline.‘”? I therefore drafted some wording to respond to paragraph
38 of the Amended GPOC in the absence of the final report (since this
had not been finalised pending receipt of Deloitte's report), which I sent
to Tony for consideration.*8° That wording was based on discussions I
had had with Deloitte.
473 After which various members of the Steering Group reviewed the draft and provided comments,
which I responded substantively to where necessary, see for example WBONO001158;
WBON0000498; WBON0000500.
474 WBONO001150; WBON0001153.
475 WBONO001154; POL00110670.
476 WWBON0000498; WBON0000499.
477 WBON0001147; WBONO0001161; POL00249903.
478 POL0024967 1; POL00249674; POL00249919.
479 WWBONO001163.
480 WBON0001164; WBONO0001165.
Page 280 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(v) I emailed Tony Robinson QC again on 13 July 2017, setting out what I
had briefed the Steering Group and General Executive about and
outlining some key points that were discussed at the meetings.‘®"
Subsequently Jane MacLeod emailed me the proposed minutes of the
General Executive meeting for my approval and confirmed that she had
read the draft (as it then stood) and approved.*®?
(vi)Owain Draper proposed his final edits on 14 July 2017.48 Tony
Robinson QC responded with some changes of his own including
(most significantly) a revised set of paragraphs responding to
paragraph 38 of the Amended GPOC, and an amendment to the
‘necessary cooperation’ term admitted at paragraph 105 of the Generic
Defence. He also sent an amended version of the suspense account
rider.4* Additionally, Amy Prime sent Owain Draper and Tony Robinson
QC an updated version of the Defence building in the comments we
had received from POL, Deloitte and Fujitsu.4®° Given the substance of
the changes Tony Robinson QC had proposed I emailed Jane
MacLeod and Mark Underwood highlighting them and explaining the
rationale.*8® Mark Underwood signalled his approval the same day*®”
(and Jane MacLeod reviewed and signed off on 18 July as set out
below).
481 POL00249919.
482 POL00024627.
483 WWBONO0000501.
484 WBON0001166; WBON0001167.
485 WWBONO001168
486VWVWBONO001171.
487 POL00024771.
Page 281 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(vii) Later on 14 July 2017, I re-circulated the draft of the Generic
Defence sent to the Steering Group and invited any last comments or
suggestions by 17 July.*8® Additionally, that afternoon Tony Robinson
QC provided an updated version of the Defence building on the version
Amy Prime had sent to him*®9.
(viii) Most of the Steering Group had no further comments on the
Defence,*® save for Rob Houghton (POL’s Chief Information Officer)
who queried two aspects of paragraphs 50(1) and 59. First, he
expressed a preference for the statement, “For a system of Horizon’s
scale, Post Office would characterise the number of errors or bugs in
Horizon requiring fixes as relatively low” to be removed. He felt that the
attribution to POL was risky given that this was Fujitsu’s, rather than
POL’s, language.‘*" I therefore deleted this sentence. Second, he felt
that administrator access should not be described as a ‘functionality’ of
Horizon, and that the Defence should instead make the point that all IT
systems have this type of access. I explained that whilst he was
technically correct we had deliberately avoided the approach he
suggested, as it was likely to be regarded as overly technical and
semantic. I slightly tweaked the wording of paragraph 59 to better
reflect Rob Houghton’s concern (“it is admitted that, although Horizon
488 WWBONO0001173.
489 WBONO0001176.
490 A few emailed to endorse it: WBON0001178; POL00024489.
491 WBON0000502.
Page 282 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
classified_as ‘functionality’ in_an_IT system, there wasis a highly
theoretical ...”), and ran both changes past Tony Robinson QC.49?
(ix)With reference to Q75.6, the contents of the Generic Defence were
derived from a combination of sources on whom it was necessary to
rely — principally POL, but also in some instances third parties such as
Fujitsu. The detailed factual investigations together with the review and
sign-off process described above were therefore designed to gather,
test and review the facts obtained from these sources. I relied upon
this process in order to be satisfied (and so to enable POL to satisfy
itself) that the contents of the Generic Defence were true. I sent the
finalised version of the Generic Defence to Jane MacLeod to review
and sign off on.4% She signed the Statement of Truth on 18 July 2017
and the Generic Defence was returned to me for filing and service the
same day.4%
482. Against the background of this overall process, I turn to the specific paragraphs
of the Generic Defence about which the Inquiry has asked and the factual
investigations undertaken that led to those paragraphs.
Investigations undertaken — remote access
483. After the preliminary report in July 2016, the focus for Deloitte was not on whether
remote access functionality was possible (which was established by that point),
but what controls were in place to minimise associated risks. This reflected Mark
Westbrook’s view that it was normal for an IT system like Horizon to have users
492 POL00024253; WBONO001180.
483 WBONO001183.
494 WBONO001185.
Page 283 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
with administrator access,**> the question being what controls were in place to
regulate this. The investigations which resulted from Deloitte’s findings about
privileged user access focused on better understanding of these controls, and,
prior to service of the Defence, were broadly as set out below. At the outset I note
that Jonny Gribben principally led on this work for WBD. I monitored progress
and was sometimes copied into emails, so although I was aware of the main
developments I was not especially close to the detail.
484. Deloitte continued their investigations into remote access and produced a report
titled ‘Bramble Draft Report: Draft for Discussion’ dated 7 October 2016 (the
“Project Bramble October 2016 Report” or the “October 2016 Report’).4%°
485. The October 2016 Report found:
485.1.In relation to the Balancing Transaction functionality, “Any writes by the
[SSC] to the [BRDB] must be audited”. The default position was that “SSC
will have privileges of only inserting balancing / correcting transactions to
relevant tables in the database”.
485.2.However, the report also stated that there was an exception in relation to
the latter point, in that “fa] small number of users are granted extended
privileges which enable them to update / delete records’.
485.3.In other words, the October 2016 Report confirmed the point that had
earlier been flagged in Deloitte's July 2016 preliminary report, namely that
SSC users with privileged access could edit and delete transactions (i.e.
not just make insertions) to the BRDB. However, the October 2016 Report
485 POL00408699.
496 WBON0000468; WBON0000469.
Page 284 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
went on to say that “the control is operating in line with management's
expectations. Access to the privileged role is restricted to users explicitly
authorised for this access. User actions are audit logged, and not
proactively reviewed’.
485.4.Jonny Gribben followed up on this, asking “what does ‘in line with
management's expectations’ mean?” And “what does not ‘proactively
reviewed’ mean and would you expect this access to be proactively
reviewed?” .49”
486. From the discussions that followed it appeared that: (i) it would be extremely
difficult for a superuser at Fujitsu to manipulate transaction data in the BRDB
without detection, as they would have to ‘fake’ the digital signature associated
with the altered transaction within a narrow temporal window (of a maximum of
15 minutes) before the relevant data was ‘collected’ and recorded to the Audit
Store; (ii) the superuser would need to write a bespoke and sophisticated
computer programme to achieve this; (iii) otherwise, there would be readily
identifiable differences between the data recorded in the Audit Store and the
(altered) data on the BRDB. Aside from it not being practically feasible for
superusers to edit or delete transaction data in the BRDB in such a way that this
would ‘track’ through to the Audit Store, there were separate audit logs recording
superuser access.
487. As to the later point, on 8 November 2016 Mark Westbrook (Deloitte) said, first:
“If you can’t fake a digital signature then for counter initiated transactions you are
unable to disguise the fact you have tampered with the data even if you edit audit
487 WBON0001064; POL00408731.
Page 285 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
logs etc”. And second: “As articulated earlier we haven't really affirmed either
way whether [superusers] can amend activity / audit logs FJ attest they can't
however’ 496
488. Investigations into this aspect in particular continued into 2017. On 11 May 2017,
Jonny Gribben copied me into an email following up on a call between him, Mark
Underwood, Deloitte and Fujitsu. He asked Mark Westbrook to (i) “produce a full
description of what a Super-User would need to do in order to amend a branch's
accounts in a way that could would [sic] not leave behind a footprint of their
activity (noting that they would never be able to completely cover their tracks
because the deletion of Super-User audit files would also leave a footprint)”.4°°
489. As I understood the position at that time, it was not possible for a privileged user
to tamper with data in any way without leaving a trace because it was not possible
to switch off the privileged user audit log without breaking the Horizon system.
This was a point on which Mark Underwood sought further information from
Fujitsu on in June 2017.5 Torstein Godeseth responded that, because of the
way in which the BRDB was configured at that time, in order for a superuser to
tamper with data without leaving an audit trail leading back to them, “it would be
necessary to take the database down and then bring it up again for the
configuration change to take effect”. Prior to the reconfiguration, it would have
been possible to switch off the privileged user audit log and switch it back on
again without needing to take the database down, but this would still leave a
record of the audit trail having been turned off and back on again, which meant
498 WBON0000472, commenting on a draft summary of the Project Bramble October 2016 Report
prepared by Jonny Gribben: POL00029104.
499 WWBONO0001078.
500 WBON0000483.
Page 286 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(in effect) that there would still be an audit trail. Torstein Godeseth closed by
“reiteratfing his] view that there is no evidence that anyone has ever actually
manipulated any audit records.” °°"
490. Whilst our exchanges with Deloitte in November 2016 and thereafter had
reassured us that it would be extremely difficult for a superuser to manipulate
transaction data in the BRDB without detection, as it was not practically possible
to ‘fake’ the digital signature associated with the altered transaction without this
being picked up in the audit data, they also informed us that non-counter
transactions (e.g. Transaction Acknowledgments posted by POL) were less well-
protected within the BRDB. Jonny Gribben and I wanted to ensure that non-
counter transactions were tested to see whether there was protection against
tampering and whether those controls were actually used in practice.
Consequently, in January 2017, Deloitte were commissioned to perform an
additional piece of work to investigate the vulnerability of non-counter
transactions to tampering.5°? Subsequently (in June 2017), Deloitte provided
WBD with a draft memo on their investigations in relation to non-counter
transactions.®° In short, this work reassured us that there were mechanisms in
place that would reveal when BRDB data in relation to a non-counter transaction
was interfered with (albeit that these controls were not precisely the same as
those that applied to transactions on the BRDB that were input at the counter).
Paragraphs 57(4) of the Generic Defence
501 WBON0000483.
502 WWBONO001070.
503 WWBON0001109; POL00031516.
Page 287 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
491. Q75.5 asks me to explain the basis on which POL pleaded the following, at
paragraph 57(4) of the Generic Defence: “To have abused those rights so as to
alter branch transaction data and conceal that this has happened would be an
extraordinarily difficult thing to do, involving complex steps (including the writing
of sophisticated computer programmes and circumvention of sophisticated
contro! measures) which would require months of planning and an exceptional
level of technical expertise. Post Office has never consented to the use of
privileged user rights to alter branch data and, to the best of its information and
belief, these rights have never been used for this purpose’.
492. I have dealt with the investigations by Deloitte into privileged user access,
together with the enquiries made of Fujitsu in the course of those investigations,
above at §§483-490. Although I did not settle the wording of paragraph 57(4) of
the Generic Defence, my view is that that paragraph substantively reflected what
we understood at the time (and communicated to the Counsel team).5°*
493. Further, when the first full draft of the Generic Defence was produced by the
Counsel team on 4 July 2017, the sections on Horizon were (as noted above)
sent to both Fujitsu®° and Deloitte®°° for review. These extracts included what
became paragraph 57(4), with Counsel’s comments in yellow highlight:
“There are a small number of Fujitsu specialists who have certain
privileged user access rights which they could in theory use to amend or
delete the transaction data for a branch. The intended purpose of
privileged user rights is system support, not the alteration of branch
transaction data. To have abused those rights so as to alter branch
transaction data and conceal that this has happened would be an
extraordinarily difficult thing to do, involving complex steps (including the
504 See the executive summary provided to Counsel (inclusive of Deloitte’s comments thereon) on
19 June 2017: WBON0001 113; POLO0174660.
505 WWBON0000491; WBON0000492.
506 WBONO001128.
Page 288 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
writing of sophisticated computer programmes and circumvention of
sophisticated control measures) which would require months of planning
and an exceptional level of technical expertise. [WERE FORMAL
PROCEDURES IN PLACE FORBIDDING THE USE OF PRIVILEGED
USER RIGHTS (OR PRIVILEGED USER RIGHTS TO CHANGE
BRANCH DATA) WITHOUT SPECIFIC AUTHORISATION, FROM
SPECIFIED PEOPLE AT FUJITSU AND/OR AT POST OFFICE (AND IF
SO) WHO WERE THOSE PEOPLE?)). Post Office has never consented
to the use of privileged user rights to alter branch data and, to the best
of its information and belief, these rights have never been used for this
purpose.”
494. As noted above, Deloitte responded with comments on various paragraphs of
the pleading, but suggested no amendments to this paragraph.” In Fujitsu’s
case, there was a call with Pete Newsome and Torstein Godeseth on 7 July 2017
which resulted in some amendments to the text (but not to this paragraph) which
Fujitsu were asked to consider further.5°® At no point did Fujitsu suggest any
changes to this paragraph, and indeed in their final response they commented:
“all fine”, subject to a couple of minor tweaks to different paragraphs of the text.5°9
The only amendment in the final version of the Generic Defence was that the
highlighted drafting comment was removed.
Paragraphs 48(3)(b) and (c) of the Generic Defence
495. Q75.3 and Q75.4 ask me to explain: (i) the “basis on which POL denied that
Fujitsu ‘edited or deleted specific items of transaction data”, at paragraph
48(3)(b) of the Generic Defence; and (ii) the “basis on which POL pleaded that
Fujitsu had not implemented fixes that had affected the reliability of accounting
balances, statements or reports”, at paragraph 48(3)(c).
507 WBON0001154; POL00110670.
508 WWBONO001137.
508 WBON0001161; POL00249903.
Page 289 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
496. It should be borne in mind, first, that paragraphs 47 to 48 of the Generic Defence
responded to paragraphs 20 to 21 of the Amended GPOC. As I understand it,
these paragraphs referred to the contractual arrangements between POL and
Fujitsu for the supply of IT services related to the Horizon system (the “POL-
Fujitsu contract”) and sought to define the scope of Fujitsu’s role pursuant to
those arrangements. Thus, paragraphs 20 to 21 of the Amended GPOC
relevantly pleaded:
“20. The Defendant entered into a contract with Fujitsu Service Limited
on 28 July 1999 for the provision of IT services relating to Horizon.
[The Claimants then referred to the fact that only a redacted version
of the contract in force since 31 March 2016 had been provided.]
21. Pending full disclosure, the Claimants understand that Fujitsu's role
included:
21.1. providing the data transfer service by which transactional data
was transferred between branches and the central data
centres;
21.2. providing a data transfer service between the central data
centres and clients of the Defendant ...
21.3. managing coding errors, bugs, and fixes so as to prevent,
manage or seek to correct apparent discrepancies in the data
(including between the said systems), in_a manner which
would potentially affect the reliability of accounting balances,
statements or other reports produced by Horizon; and
21.4. providing a telephone advice service, for and on behalf of the
Defendant ...”
497. Paragraph 21.3 of the Amended GPOC, in the wider context of paragraphs 20
and 21, was (and is) not easy to understand. Read literally, it appears to allege
that the POL-Fujitsu contract contemplated that it would be part of Fujitsu’s “ro/e”
to implement fixes “in a manner which would potentially affect the reliability of
accounting balances”. POL certainly did not accept that Fujitsu was contractually
entitled (still less obliged) to take action which would “affect the reliability’ of
Page 290 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
accounting data in this way. At the same time, unlike the rest of paragraph 21,
paragraph 21.3 appears to allude to what the Claimants believed Fujitsu to be
doing in practice, i.e. regardless of what the POL-Fujitsu contract said.
498. Whilst I did not settle the wording of paragraph 48(3) (this being Counsel's job),
when that paragraph is viewed in the above context it seems to me that its
purpose is to concisely identify the boundaries of Fujitsu’s legitimate role under
the POL-Fujitsu contract, albeit against the backdrop of the Claimants’ wider
factual allegations. Understood in this way, paragraph 48(3)(a) accepts that it
was part of Fujitsu’s “role” to identify and remedy bugs in Horizon, but
subparagraphs (b) and (c) deny that this extended to remedying errors by
“editling] or delet[ingf transaction data, or by implementing fixes which
compromised the “reliability” of accounting data:
“48(3) Paragraph 21.3 bundles together several different concepts and
uses language that is open to different meanings and/or misleading.
However:
(a) Fujitsu's role included identifying and remedying coding errors
and bugs in Horizon.
(b) To the extent that the phrase ‘correct apparent discrepancies
in the data’ is intended to mean that Fujitsu implemented fixes
that edited or deleted specific items of transaction data, that is
denied.
(c) It is denied that Fujitsu has implemented fixes that have
affected the reliability of accounting balances, statements or
reports.”
499. In any event, and to the extent that paragraphs 48(3)(a) and (b) are read as a
denial in fact that Fujitsu had ever (i) edited or deleted transaction data in order
to implement a fix or (ii) implemented a fix that impacted the reliability of branch
accounting data, it is important to stress that this wording was submitted to
Page 291 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Fujitsu for approval on 11 July 2017. Fujitsu responded suggesting no
amendments to this wording, saying “all fine apart from a couple of minor
amendments” which they made to different paragraphs in the draft.°1° How
Fujitsu had used their capabilities in practice was, of course, a matter peculiarly
within their knowledge (and Deloitte had not so far found any evidence that
Fujtisu had actually edited or deleted transaction data (as distinct from the one
then-known injection of a Balancing Transaction)).
Paragraphs 43(1) to (3) of the Generic Defence
500. Q75.2 refers me to paragraphs 43(1) to (3) of the Generic Defence, and asks me
to explain the “basis on which POL pleaded that ‘The blocked value is not (and
is not treated as) a debt due to Post Office”.
501. Having reviewed my firm’s file it appears that Amy Prime was principally tasked
with liaising with POL to obtain background information relevant to this
pleading.5"' Amy liaised directly with POL and I was generally not copied into
these emails. However, I did attend a meeting with Tony Robinson QC, Owain
Draper, Kathryn Alexander and Huw Williams on 7 June 2017 (see above,
§481.4(ii)) at which “the processes of ... end of trading period” was on the
agenda, although I have no recollection of the specifics of what was discussed
at this meeting other than that it was a long discussion about accounting
practices at POL that largely took the form of a question and answer session
between Counsel and the POL attendees.°"?
510 WBONO0001161; POL00249903.
511 WBONO0000474.
512 WBON0000481; WBONO001094.
Page 292 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
502. I was then copied into a short briefing note which Amy sent to Counsel on 26
June 2017, in which she explained the that: “Where an item has been settled
centrally and disputed, the agent accounting team apply a dunning block to the
open item on the account. This prevents any further requests for payment being
sent to the agent ... The effect of placing these blocks on the system is to prevent
a dunning letter (chaser letter) being produced and/or preventing the item from
falling on a processors worklist for further action” 5"3
503. Tony responded to Amy the same day (again, with me in copy) with various
follow-up questions largely about how this translated visually on the SPM’s
account."
504. On 27 June 2017, before Amy resolved these queries, Tony Robinson QC sent
me a partial first draft of the Generic Defence containing the following paragraph
which was a forerunner to paragraph 43(3):515
“Where this process discloses a shortfall and the Subpostmaster
disputes liability for the shortfall, he or she is required to raise this a
dispute by calling the Helpline referred to in paragraph [XX] below. If the
shortfall is for less than £150, he or she is required to make it good by
adding cash or a cheque to the branch pending resolution of the dispute
(on the basis that it will be repaid to the Subpostmaster if it is ultimately
determined that he or she is not liable for the shortfall). If the Shortfall
was for £150 or more, [the SPM] could settle it centrally pending
resolution of the dispute. In that situation, Post Office would put a block
or hold on the relevant debit created in his or her account with Post Office
(i.e. the amount settled centrally) until the resolution of the dispute.
Unless and until it was ultimately determined that the
Subpostmaster was liable for the Shortfall, that debit was not (and
was not treated as) a debt due to Post Office” (emphasis added).5"°
513 WBON0001098.
514 WBON0000482.
515 WBONO001119.
516 POL00249555.
Page 293 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
505. On 3 July 2017, Amy Prime responded to Tony Robinson QC’s queries by email
(again, with me in copy).5'” Around the same time, I appear to have had a call
with Angela Van Den Bogerd to discuss “the processes for SPMRs to dispute
TCs / Shortfalls’, but I cannot remember this call or if it related to this particular
paragraph.>'®
506. Later on 3 July 2017, I sent Tony Robinson QC my comments on the partial
draft.5'° I can see that I substantially cut the above paragraph down, including
removing the final sentence, as shown here:
“Where this—process—diset. a_shertfal_and the Subpostmaster
disputes liability for the shortfall, he or she is required to settle centrally
the shortfall (thereby bringing the branch accounts into balance) and
then raise this a dispute by calling the Helpline._Raising a dispute
suspends the payment of the shortfall that has been transferred to the
dding-cash-oracheaue-to-th.
'g cs
bo-ershe-is-reauiredtomakeitgqoodby
g y
braneh—pending_reselution_of _thedispute n-the-basisthatitwillb.
P 'g Pr t
id tothe Subpostmasterifitis ultimately determined that h. A,
P > x
is notliable forthe shortfall) Ifthe shortfallis for £150 or more _he-ersh.
n settle it contrally nding-resolution of the -dispute-—In that situation.
YP. F PL: 7
Post Office-put block-_on_the_relevant debit created_in_his_or-her
a t-with-Post-Officei.e—the—amount_settied. tral) —untith
507. My suggested edits therefore specifically omitted the reference to a blocked
value not being treated as a debt due to POL, although I cannot now recall the
reasons why I did this at the time.
517 WBONO0001125.
518 WBON0001124; WBON0000334.
518 WBONO0001126.
Page 294 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
508. Tony Robinson QC sent a complete and revised first draft of the Generic Defence
to myself and Owain Draper the following day, 4 July 2017.52 He had accepted
my revisions, save adding back in a version of the final sentence that I had
deleted (“The blocked value is not (and is not treated as) a debt due to Post
Office”). The relevant paragraph therefore now read as it does in the final version
of the Generic Defence.
509. Extracts of the draft Generic Defence containing this paragraph were sent to
Kathryn Alexander and Shirley Hailstones for review on 5 July 2017.52? They both
returned the draft without raising any issues with this paragraph. The relevant
sections were then sent to Angela Van Den Bogerd for a second-line review on
10 July 2017.54 Again, she returned her comments without raising any issue with
the characterisation in this paragraph.5?>
(iii) The Project Bramble Report (Q89)
510. I have referred above to the investigations by Deloitte which were fed into the
LOR and Generic Defence (§§466-472; §§483-490). These were largely
complete by the time the Defence was served, and on 1 September 2017 Deloitte
produced a draft report which represented the culmination of their work (the
“draft Project Bramble Report”). I was sent this draft report the same day;5%
this document is POL00041491 to which I am referred by Q89 of the Request.
520 WBON0000485.
521 WBONO0000486.
522 WBON0001130; WBON0001132.
523 WBON0001141; WBON0001142; WBONO0001143; WBONO001144.
524 WBON0000493; WBON0000492.
528 WBON0001150; WBON0001153.
526 WBON0001192.
Page 295 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
511. I do not describe the detailed content of the draft Project Bramble Report here
save to note that it did not contain any findings (for example, with respect to the
nature and extent of Fujitsu’s remote access capabilities) that substantially
changed my understanding of the matters Deloitte had been investigating,
compared with what I had understood to be the case at the time the Generic
Defence was finalised. That was not surprising: we had been working closely
with Deloitte for more than a year by this point so were aware of the investigations
they were pursuing and their emerging findings.
512. However, and with reference to Q89.1 and Q89.3, I considered that the language
Deloitte used in the draft report was more heavily caveated than I was expecting
based on our prior interactions and the wording they had approved in the Generic
Defence (see above, §481.5 and §§493-494). In verbal conversations they had
been clear and confident that Horizon was reliable, and I felt that this was not
properly reflected in the draft report. My understanding was that the substance
of their views had not changed, but that the draft report understated their real
level of confidence in Horizon. My impression was that this was a function of the
fact that Deloitte were, as organisation, concerned about putting down their views
in writing in case it created a risk for Deloitte; everything in writing had to go
through several levels of checks before it could be approved for release. The
other challenge we had was that Deloitte struggled to articulate their views in
language that was easy to understand and not excessively technical. The draft
Executive Summary at POL00041492 reflects what I understood to be Deloitte’s
true views, based on our work with Deloitte to date including verbal interactions
with them.
Page 296 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
513. With reference to Q89.2, I vaguely recall that I had a call with Andy Whitton of
Deloitte about this. I shared my frustrations with him but I also recall making it
clear (as I had on several calls with Mark Westbrook) that I was not looking to
push for a particular conclusion and that Deloitte should only give views they
believed. Since the Project Bramble report was intended for POL’s internal use
and was not going to be relied on as evidence, I saw no reason why Deloitte
should not produce a report that candidly and plainly stated their views (nor would
there have been any benefit for POL in having a report that did anything else).
The issue was that I understood Deloitte to believe that Horizon was reliable, but
they were unwilling or unable to reflect this fully in their draft report due to what I
perceived as institutional aversiveness to committing themselves in writing.
514. I conveyed these views to Rodric Williams in my dated 27 September 2017
(POL00041490) and explained that I did not feel that Deloitte could be stood up
as POL’s IT expert in the group litigation. To be clear, this was my view from the
outset of the litigation (as made clear for example in an email I sent to Tony
Robinson QC on 8 June 2016, see above at §420.7), because to use Deloitte as
our expert witness would risk waiving privilege over their instructions and work
product to date, and due to their historic engagements by POL, they lacked the
necessary independence to be a testifying expert in Court. The draft Project
Bramble Report therefore did not change this view, but Deloitte’s unwillingness
to reflect their real views in writing (together with their impenetrable writing style)
did underline my existing views as to why it would be unwise to instruct them as
our expert witness in the litigation. My recollection is that these same views were
shared by the Counsel team.
Page 297 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
515. With reference to Q89.4, the sentence which the Inquiry cites had already
appeared in Deloitte’s earlier preliminary report dated July 2016. Detailed
investigations into Fujitsu’s remote access capabilities were undertaken as a
result, as I have set out; the draft Project Bramble report represented the
culmination of these investigations by Deloitte rather than the beginning of them.
516. For completeness, in October 2017 a near final draft of the Project Bramble
Report was provided, in which Deloitte maintained that its testing supported the
view that neither POL or Fujitsu had the ability to log on remotely to a Horizon
terminal in a branch so as to conduct transactions nor to push transactions
without an SPMR’s knowledge or consent, with the exception that “a small group
of Fujitsu privileged users ... may do so via Balancing Transactions” 52”
517. Several rounds of comments were exchanged between POL, WBD and Deloitte
about the drafting of the report over the next couple of months, but the essence
of the conclusions did not change. The final version of the Project Bramble
Report was produced and circulated by Deloitte on 2 February 2018.57 It
concluded:
“A limited number of authorised Fujitsu personnel (19 at the Operating
System layer and 26 at the database layer at the time of testing - May
2016) have sufficient privileges to theoretically add /delete / change data
in the BRDB (‘Privileged Users”). However, see paragraph 1.4.2.10
below regarding the segregation of access conditions. These users may
also have access to other systems, such as the Audit Store, however in
relation to the allegations, access to the BRDB is the most important as
it is the BRDB that generates the branch accounts and is the source of
the data ultimately used by Post Office to investigate shortfalls."
527 WBON0001209; POL00139454.
528 WWBON0001223; POL00139537. Although still marked as ‘draft’, this was the final version to the
best of my belief.
Page 298 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
518. Deloitte also concluded that it was likely that any changes made by privileged
users would be likely to be identified and resolved. It would be such a complex
exercise for a privileged user to cover up any changes they had made that it was
unlikely in practice that they would be able to do so:
"While we have identified an exception in the cryptographic controls
(paragraph 1.4.2.10 and 1.4.2.11) which would theoretically allow a
malicious actor to undermine them and potentially change data, it is
limited to a third party (Fujitsu) and would be technically very challenging
to achieve. It would require significant motivation for one of the limited
set of Fujitsu staff members to exploit this vulnerability given the
technical challenges and risks of tripping monitoring controls and,
although we have not performed procedures in this area, it would almost
certainly require collusion with Post Office staff or Postmasters.
Although our investigations have not been exhaustive, they have been
extensive and we have seen no such evidence of malicious misuse of
the system."
519. Further, Deloitte said that, in relation to the flow of core data within Horizon Online
from counters in branch to the Audit Store, the controls in Horizon “represent the
most reliable control type possible over data integrity’. These controls, in
Deloitte’s view, made it “it extremely unlikely that the record of transactions
contained within the Audit Store is not representative of the transactions input by
Staff in branch. As with all large scale computer systems whilst it is theoretically
possible that glitches and coding errors in the system could have resulted in
errors in the recording of transactions to occur, the likelihood of such errors
occurring in a manner which has adversely affected only certain branches
materially whilst not affecting other branches at all / minimally is in our view
remote given the controls in place. The testing we have performed over these
controls was designed and executed to assess their operation in responding to
these fundamental risks. Noting the assumptions and limitations detailed in
section 1.5, this testing has not resulted in any matters being identified that would
Page 299 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
call into question the integrity of the core data flow within Horizon Online from
the Counter in branch to the Audit Store. I am clear that the views expressed
above and in the final version of the Project Bramble report were entirely
Deloitte’s own.
520. Again for completeness, although Deloitte’s work on Project Bramble was
substantially complete by the end of 2017, in 2018 WBD instructed Deloitte to
review two batches of technical documents provided by Fujitsu to identify
whether there was anything that undermined or contradicted the Project Bramble
report. My understanding was that they did not identify anything of significant
concern.522
N. DISCLOSURE (Q58.4, Q76 to Q87, Q88.3, Q90.1, Q91, Q95.1, Q99)
(i) Introduction
521. This section addresses the Inquiry’s main questions about the advice IMWWBD
gave POL in relation to the disclosure and redaction of documents, namely:
521.1.Q58.4 (general advice on disclosure);
521.2.Q76 to Q82 (disclosure of the KEL database);
521.3.Q81.2 and 83 to Q87 (disclosure of the ‘Peak’ database);
521.4.Q88.3 (disclosure of the reports generated by Project Zebra);
521.5.Q90.1, Q91 and Q95.1 (approach to redacting evidence deployed in the
CIT and HIT); and
528 WBON0001248; POL00028982.
Page 300 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
521.6.Q99 (obtaining documents held by Royal Mail).
522. There are other questions related to the broad theme of disclosure which I
address elsewhere as appropriate. For example, the Inquiry’s questions relating
to preservation of documents and POL’s response to particular early requests for
documents by the Claimants are dealt with above in Section L. I address the
discovery of back-versions of KELs in October 2019 in Section R below, since it
relates to events post-dating the HIT.
(ii) General advice on disclosure
Factors relevant to disclosure generally
523. The disclosure exercise — or more accurately exercises — in this case were an
enormous undertaking. Thought was given to how disclosure ought to be
managed from an early stage, and it is difficult to overstate the range and
complexity of the factors in play. By way of brief summary only:
523.1.Given the size of POL's business and the scale of the Horizon system, the
pool of potentially relevant documents was vast.
523.2.The claims in the litigation spanned a period of almost two decades.
523.3.Many potentially relevant documents were of a highly technical nature.
523.4.Some of those documents were in the hands of other parties such as Royal
Mail, Fujitsu and ATOS.
523.5.There were a wide range of different formats in which potentially relevant
documents existed, including more difficult to disclose formats such as
databases.
Page 301 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
523.6.POL’s IT systems on which they stored those documents were particularly
complex. They had also changed substantially over the years, including
notably when it separated from Royal Mail.
523.7.Different parts of POL stored documents in different ways, subject to
different retention policies, and there was very little top-down
understanding within POL of what the potentially relevant documents were
or where they were. Many classes of potentially relevant document — for
example, documents relating to branch audits and investigations — were
held by multiple teams.
523.8.In the LOC and indeed the Claimants’ generic pleadings (served a year
later), I believe it is fair to say that the claims were very wide and not clearly
articulated. This presented a real challenge in terms of understanding the
issues and what documents were necessary and proportionate to disclose,
and when.
523.9.There was a sizeable and growing number of Claimants until the eventual
GLO cut-off date in December 2017. At that stage there were around 550
Claimants, some of whose claims dated back many years.
523.10. Other than the limited information in the SOls and information that some
Claimants had provided during the Mediation Scheme, POL had little detail
on the specific facts and matters disputed by the vast majority of the
Claimants (setting aside the six Lead Claimants for the CIT, in respect of
whom individual particulars of claim were produced).
Page 302 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
523.11. Since the subject matter of the group litigation was still live (including
that some Claimants were still in-post as SPMs), there was the possibility
that new potentially relevant documents were being generated all the time.
524. Given the above challenges POL was understandably concerned by the
prospective costs of the disclosure exercises in the group litigation and was
anxious that they remain proportionate. But this in turn was difficult to assess
because there was much uncertainty around the true value of the Claimants’
claims.
525. At the same time, this was a case in which the Claimants were alleging
systematic concealment and deceit and POL, on our advice, recognised that it
had made incorrect statements in relation to remote access. Accordingly,
Counsel and I had advised, and POL understood, that failure to disclose
documents would feed the Claimants’ narrative around concealment (which as
explained above was of central importance to the key issues of limitation and
past settlements).
The process and sequencing of disclosure in the group litigation
526. As noted above, Mr Justice Fraser ordered a multi-stage trial process, with
disclosure being ordered in tranches that mirrored the case management
directions and trial sequence; i.e. disclosure for the CIT started to be given first,
then disclosure for the HIT (with an overlap between the relevant disclosure
exercises), and so on. Further, the parties and the Court agreed to follow the
newly developed Disclosure Pilot under the CPR (which has since been fully
adopted). The core objective of the pilot was to reduce the cost of disclosure by
reducing the scale of disclosure exercises. It actively discouraged the old model
Page 303 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
of ‘standard disclosure’ and promoted disclosure orders that targeted narrower
classes of document. This was the approach adopted by Mr Justice Fraser and,
accordingly, at no stage was POL subject to a general order to disclose all
relevant documents.
527. The combination of staged trials and the approach adopted under the Disclosure
Pilot rules had a number of consequences for the disclosure process in this case:
527.1.First, it led to disclosure being ordered at different CMCs as the litigation
progressed, with the relevant directions being split across several different
Orders starting from October 2017. The Orders were typically prescriptive
as to the disclosure to be given, often including carefully worded schedules
of classes of documents that were either negotiated between the parties
or determined by the Court.
527.2.Second, the fact that the first two key trials — the CIT and the HIT — were
listed to follow each other in quick succession made for a challenging
process with disclosure being given on a rolling (and sometimes
simultaneous) basis in preparation for each of the trials.
527.3.Third, with each of the trials being intended to resolve groups or ‘buckets’
of cross-cutting issues (as opposed to, for example, a more traditional
unitary test case trial), there were difficult questions of whether disclosure
was inside the scope of a disclosure order or not.
527.4.Fourth, since the issues in dispute relative to different trials were
ventilated at different points in time, the relevance of different documents
to those issues only became known at different times. Most important in
my mind was that the Claimant's case in relation to Horizon was sparsely
Page 304 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
explained in the LOC and Amended GPOC, and did not start to be
revealed until mid-2018 (when the Claimants were, for example, required
to produce an outline document setting out their allegations in relation to
the Horizon system; see above, §321.4).
528. Reflecting back now, I can see how this approach contributed to some of the
problems with POL's disclosure. Disclosure by way of narrow classes of
documents meant that sometimes material documents were disclosed later than
was ideal because they fell outside the scope of the disclosure ordered. To give
one example — the Court never ordered, and the Claimants never sought,
disclosure of the Peak database and so it did not form part of the disclosure for
the HIT. It was later voluntarily disclosed by POL on 27 September 2018 after the
experts indicated in July / August 2018 that they thought that the documents
contained within the Peak database were important.
Overview of disclosure advice
529. I and my colleagues at WBD were conscious of the above considerations. I would
describe our overall approach as balanced and striving for proportionality —
bearing in mind the risk of high disclosure costs and not being able to complete
disclosure within the timescales set by the case management Orders in the group
litigation — whilst erring on the side of recommending that access to documents
be given where reasonably possible. Whilst, as I shall presently explain, there
were certainly challenges in relation to the disclosure process, we acted at all
times in good faith — and in my view, we not infrequently recommended
disclosure above and beyond what was required under the formal disclosure
orders. I would add that Counsel’s advice was taken on important aspects of
Page 305 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
disclosure and so the advice given to POL on these questions was very much a
team effort.
530. With reference to Q58.4, it is difficult to be much more specific than this in the
present context. The sheer volume of emails and other documents which my firm
has on file relating to these matters (and the number of those that are likely to
relate to points of detail on disclosure) is such that it has not been possible in the
time available to sensibly search and review all of this material for the purposes
of preparing this statement. I have however reviewed the Decision Papers which
WBD presented to the Steering Group over the course of the group litigation, and
I set out below a summary of some of the main decision points and
recommendations, in chronological order and in the context of the Orders made
at the CMCs, to give a flavour of the overall advice we gave. The Decision Papers
referenced below should be read together with those concerning preservation of
documents, which I have summarised above at §399.
531. By a Steering Group paper dated 5 October 2016 (after service of the first Claim
Form but before the Group Litigation Order was made), WBD advised that POL
should allow Freeths access to Second Sight to discuss the Claimants’ cases
pursuant to a protocol to be agreed between the parties so to protect POL's
privileged material that Second Sight had historically had access to.°°
532. By a Steering Group Paper dated 14 February 2017 (after the Group Litigation
Order was made but before any disclosure was ordered), WBD advised that POL
give further voluntary disclosure to Freeths:
“Disclosure:
530 POL00139321.
Page 306 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Further disclosure should be voluntarily given to Freeths for the following
reasons:
a. Giving disclosure on the above Target Issues should make it easier
for a Court to select those issues at the CMC.
b. Freeths are hiding behind the lack of disclosure to avoid explaining
weak points in their case. The lack of clarity allows them to keep
unmeritorious claims alive.
c. We believe that Freeths are struggling to cope with the volume of work
and extra disclosure makes their job more difficult. Work pressure drives
a greater prospect of a good settlement. The areas where further
disclosure could be given are set out in Schedule 1.”5°"
533. Accordingly, we were advising in favour of voluntary disclosure for three
reasons. First and foremost, it would be helpful to the Court to identify the key
issues (point (a)). Secondly, it would assist our efforts to press the Claimants to
particularise their cases better and so identify cases which lacked merit (point
(b)). Thirdly, we saw a benefit in terms of litigation tactics, in that disclosure
would increase the burden of work on Freeths, which might in turn make them
more amenable to consider settlement (point (c)). I acknowledge that the third of
those reasons was a matter of tactics, but in adversarial litigation it is sometimes
necessary to give such advice to clients. It will also be clear from the paper that
we considered the disclosure proposal to be procedurally right as well as having
potential tactical benefits.
534. The issue of allowing Freeths access to Second Sight was returned to in a paper
dated 12 July 2017, a protocol by now having been negotiated between WBD,
Freeths and Second Sight. We advised that POL should now allow Freeths
access to Second Sight pursuant to that protocol, noting that not to do so would
531 POL00247209.
Page 307 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
likely be regarded as unreasonable by the Court.5°? The protocol permitted
Second Sight to discuss a range of key topics with Freeths, including (but not
limited to): the architecture of the Horizon system; the installation and
implementation of Old Horizon and Horizon Online; the differences between and
capabilities of the two; updates and software versions since installation;
Transaction Corrections; the NSBC Helpline and the technical helpline operated
by Fujitsu; problems with hardware; errors, bugs, fixes, issues and ‘Peaks’; the
KEL; the “extent of error repellency in the Horizon system”; and POL’s access to
transaction data and its agreement with Fujitsu in respect of provision of such
information 53
535. The formulation of orders for disclosure came into focus in advance of the first
CMC on 19 October 2017 through exchanges of correspondence between the
parties. In October 2017, WBD explained by way of update to the Steering Group
that:554
535.1.Freeths were seeking a very substantial disclosure exercise in advance of
any trial, the first of which should deal with only a narrower range of
contractual issues than POL thought should be considered (with Freeths'
proposed issues excluding, for example, any issues concerning the NTC).
Freeths’ essential position was that POL had been obstructive in the past
and held most of the relevant information, and that that imbalance should
therefore be rectified in advance of any trial.
532 POL00139406.
583 POLO00250171.
534 POL00006431.
Page 308 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
535.2.WBD did not agree with that assessment. The disclosure sought by
Freeths was massive, untargeted, and could cost up to £7m more than
staging disclosure by reference to each trial — without any obvious benefit.
It would also likely make it impossible for the first trial to be held before
late 2019 or early 2020.
535.3.In relation to disclosure of the KEL, POL had offered the Claimants’ IT
expert direct access to the KEL at Fujitsu’s Bracknell office, but Freeths
maintained that disclosure of the whole KEL database should be given.
WBD proposed to maintain POL’s position at the CMC; I explain the
reasons for this further below at §§553-554 and §594.
536. As noted above (§§317-318), at the first CMC in October 2017 the Court set the
Common Issues down for trial and in summary made the following orders in
relation to disclosure:
536.1.The parties were required to give disclosure of prescribed classes of
document in relation to each of the Lead Claimants which they selected
for the CIT.
536.2.POL was to disclose a small number of technical documents in relation to
Horizon and documents originally held by Second Sight.
536.3.The parties were to each produce an EDQ so that further orders for
disclosure could be made at a subsequent CMC.
537. In December 2017, WBD provided the Steering Group with an update on the
EDQ process. WBD advised that:
“Relevant individuals and teams filled out questionnaires in relation to
the matters covered by the EDQ ... Since then, we have covered this
Page 309 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
ground again with the business, holding calls with each business area,
Fujitsu and Post Office IT teams. This has been to double check our
understanding and to make sure we pick up less important locations of
documents that were not needed for the CMC witness statement ... This
process has enabled us to provide a generous amount of information
within the EDQ in order to assist (and be seen to be assisting) the Court
in making further directions as to disclosure ... it is important that Post
Office is seen to be assisting the Court in understanding the vast extent
of the documents it holds”.5*
We further advised that there was a “wider strategic question about the purpose
for which disclosure is being given”, on which advice was being sought from
Counsel. In short, we anticipated that Freeths would make wide-ranging requests
for disclosure based on the detailed information provided in the EDQ, and that in
order to keep the scope of disclosure within manageable and proportionate
bounds, it would be necessary to provide the court with a reasonable alternative
scope of disclosure.
538. A Steering Group paper dated 4 January 2018°°° recorded that the parties’
lawyers had had a “cordial and constructive” meeting ahead of the CMC listed
for 2 February 2018 to discuss the future direction of the litigation and, in
particular, disclosure.5°”7 However, we had formed the impression from the
meeting that “Freeths have done little forward planning beyond November 2018
[when the CIT was listed to take place] and they do not really understand how to
undertake a major disclosure exercise.” They were maintaining, in effect, that
wide-ranging disclosure should be given on all issues ahead of the first trial,
which we considered to be wholly unfeasible and could derail the litigation
timetable. Our position therefore (which we recommended be maintained) was
535 POL00357949.
536 NB the date 4 January 2017 is incorrectly given on the document.
537 POL00252428.
Page 310 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
that we should seek staged disclosure on a trial-by-trial basis, which would still
give the Claimants access to the documents needed to resolve the issues at
each given trial. Given the differences between the parties we considered it “likely
that we will be going into the 2 February 2018 hearing without any agreement on
disclosure”. We also anticipated that at some stage POL would need to disclose
documents from a system it used called SharePoint. We recommended that
steps be taken to extract the whole of SharePoint immediately so that (i) the
disclosure arising out of SharePoint could be scoped; (ii) early disclosure of
relevant documents could start to be given to the Claimants; and (iii) inadvertent
deletion could be avoided. We made this recommendation notwithstanding that
extracting the whole of SharePoint was likely to be a costly exercise (in the region
of £145,000 with monthly hosting costs of around £15,000).°°
539. On 31 January 2018, Amy Prime sent Mark Underwood a Noting Paper titled
‘Update on strategy for the Court hearing on 2 February 2018’ (being the second
CMC).%° In the Noting Paper, WBD advised the Steering Group that the parties
had narrowed their differences on disclosure, with the Claimants now accepting
that the disclosure should be staged by trial. However, there remained a dispute
as to the Disclosure Pilot ‘Model’ to be adopted in relation to generic (as opposed
to Claimant-specific) documents required for the CIT, with the Claimants seeking
Model D search-based disclosure which would result in (we felt) the disclosure
of an excessively wide range of documents which did not go to the contractual
relations between the parties. We advised that POL should seek Model C narrow
538 Over the following months we conducted reviews of POL’s c.500 SharePoint sites to identify
which should be extracted in full. We kept Freeths informed of our decision-making in relation
to which sites should, and which need not, be extracted, but received no response. This was
set out in an update provided to the Steering Group on 11 April 2018: POL00254458.
538 POL00253188; POL00139539.
Page 311 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
class-based disclosure, by reference to around 30 classes of documents. The
reason for our advice was that it was an important consideration that the costs of
the document review exercise be kept proportionate and that the trial timetable
be maintained; we were concerned that the Claimants’ approach would result in
millions (as opposed to hundreds of thousands) of documents being disclosed,
which would be difficult to achieve before the end of 2018.
540. At the second CMC on 2 February 2018 (referred to at §320 above), the Court
made the following orders™° relevant to disclosure:
540.1.The Court agreed with POL's general approach that disclosure should be
given in stages, broadly aligned to the scope of each trial. I recall there
being a discussion before Mr Justice Fraser about the scope of the
disclosure orders for the CIT needing to be limited to the admissible factual
matrix for the SPMC and NTC.
540.2.Disclosure would be in accordance with Model C (narrow classes of
documents), being POL's preferred approach as opposed to the wider
issues-based approach to disclosure advocated by the Claimants.
540.3.It was directed that the parties were only required to undertake a
reasonable and proportionate (not exhaustive) search for documents
within any class.
540.4. There would be disclosure of documents in relation to each Lead Claimant
for the CIT in line with the search and keyword criteria set out in the
540 WBON0001230.
Page 312 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
schedules to the Court's Order ("Stage 1 Disclosure") by 28 February
2018.
540.5.There would be further disclosure of generic documents for the CIT
("Stage 2 Disclosure") by 18 May 2018. This was subject to the parties
agreeing narrow classes of documents and a list of custodians of
documents at POL against whom searches would be run to locate those
classes of documents.
541. The Court having set down the principles and structure for disclosure in the
litigation, a third CMC was then listed for 22 February 2018 to determine the
exact narrow classes of documents to be disclosed as part of Stage 2 Disclosure.
I emailed Rodric Williams and Mark Underwood on 12 February 2018 setting out
our advice on the approach to take at the next CMC.
"We have reviewed the Claimants' requests for Model C disclosure.
They are, in effect, still seeking massively wide disclosure that goes far
beyond the Common Issues for November and far beyond admissible
factual matrix. The attached Model C table includes our comments on
each request (which has a few points in yellow that need finalising
tomorrow).
We recommend that Post Office opposes nearly all these requests, save
for those that are sufficiently narrowly defined that giving them would be
easy. We believe that it is important that Post Office adopts a consistent
approach. If we oppose certain categories of documents on the grounds
that they are inadmissible, then we need to oppose all similar documents
save where there is an obvious reason not to do so — which leads to a
large number of requests being opposed. We should also keep in mind
that Post Office's original Model C proposal was drafted very generously
and in places went beyond admissible factual matrix, so the Cs are
already getting more than they are strictly entitled to.
Page 313 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Counsel and I have a fair degree of confidence that the Court will be with
us on this approach so long as we continue to constructively engage with
Freeths. "©"
542. As this was a continuation of the same issues that had been raised with the
Steering Group on 31 January 2018, I did not believe that this approach needed
a formal Steering Group Decision Paper. Mark Underwood agreed with that
assessment? and I believe that I verbally briefed the Steering Group about this
approach on 14 February 2018 (though I do not recall this meeting).>4? Before
the 22 February 2018 CMC the parties were able to agree a list of 51 custodians
at POL whose accounts would be searched for disclosable material, and
although some progress was made in agreeing the classes of documents many
were still in dispute — largely on the grounds that the Claimants were either
seeking inadmissible material or the request did not amount to a narrow class.°4
543. At the third CMC on 22 February 2018, the following relevant directions were
given (see §321 above):*4>
543.1.The classes of document for Stage 2 Disclosure for the CIT were ordered.
My recollection is that the Court agreed more with POL's proposed
disclosure orders than the Claimants’ proposals.
543.2.By 19 April 2018 the parties and their experts were to meet and attempt to
agree (i) the scope of any further information or documents relating to
Horizon that the experts required, and (ii) a process for the experts to
inspect the Horizon system.
541 POL00253355.
542 WBON0001226.
543 POL00253363.
544 WWBON0001229; POL00408810; POL00253516.
545 WBON0001232.
Page 314 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
544. The parties’ solicitors and their experts met as ordered on 11 April 2018.
Following that meeting Freeths made requests for (i) further inspection of
Fujitsu's systems, (ii) further information about Horizon, and (iii) further disclosure
of documents. I emailed Jane MacLeod, Tom Moran, Rodric Williams and Mark
Underwood for instructions ahead of a CMC to consider these matters that was
scheduled for 19 April 2018 (although in the event this was vacated).™° I set out
the broad range of options as follows:
POL's options are:
1. Refuse all requests for documents / information. Not
recommended — its overwhelming clear that further information on
Horizon is needed, it’s a question of how much and how to deliver it.
2. Seek to adjourn the CMC — Not recommended — our judge does
not like delay and this will look obstructive.
3. Provide what is reasonable - Recommended — POL provides
what it considers reasonable in the circumstances and stands it [sic]
grounds on the unreasonable requests.
545. Specifically in relation to disclosure, I noted that the Claimants had made a "Wide
requests for lots of documents" and recommended to POL that it should "Re-
draft the requests so that they are narrower following the strategy adopted at
previous CMCs and then agree to give those narrower classes of document. Only
oppose requests that are plainly unreasonable." POL approved this approach®4”
and WBD wrote to Freeths as instructed.548
546. Through negotiation and agreement between the parties, some classes of
documents were agreed for disclosure. However, others remained in dispute and
546 POL00022706. This email refers to an imminent CMC in April, which was then adjourned to 5
June 2018.
547 WBON0000177 and POL00022705.
548 POL00254578.
Page 315 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
were determined by the Court at a further CMC on 5 June 2018 (referred to at
§322 above). The final list of classes of documents to be disclosed by 17 July
2018 were recorded as “Stage 3 Disclosure” in the schedule to the Order
following this CMC.*4° This Order also provided for inspection of Fujitsu’s Peak
and TFS systems by the parties’ IT experts to be facilitated, and for the experts
to submit any requests for further information (not documents) about Horizon by
26 July 2018.
547. From this point onwards, there were no more disclosure orders made by the
Court in relation to either CIT or the HIT (putting aside issues relating to particular
documents that arose during the course of either trial).°5° Further requests were
made through correspondence, each of which WBD reviewed on its merits. In
relation to such further requests by the Claimants, WBD advised POL to agree
to requests where it was reasonable and proportionate to do so.
548. An example of this is the Steering Group paper dated 26 September 2018, by
which WBD advised POL to disclose the Peak system following requests by the
Claimants’ expert for certain categories of Peak (as to which see further below,
§§621 ff). The reasons for our advice were: (i) providing disclosure of this scale
voluntarily would be viewed favourably by the Managing Judge; (ii) it reflected
and continued POL’s intent to provide assistance to the Claimants where it was
reasonable and proportionate to do so; (iii) it should neutralise some of the
Claimants’ expert's requests; (iv) it would assist POL’s own IT expert, who could
548 POL00120352.
580 We also sent detailed letters to Freeths explaining the approach we had taken to give the
ordered disclosure so that they might challenge that approach if they disagreed with it, see for
example POL00285777; WBON0001690; POL00285778. In my view, this provided a level of
information about disclosure that went above and beyond what would normally be required in
civil litigation.
Page 316 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
not be provided with the Peaks unless they were also made available to the
Claimants’ expert; and (v) it is likely that certain of the Peaks were adverse and
would be required to be disclosed at some stage in any event.°°"
549. Beyond this there were around 12 further tranches of documents disclosed
between August 2018 and the start of the HIT in response to requests that were
made by the Claimants as the litigation progressed or as WBD identified adverse
documents that POL was required to disclose in the course of preparing evidence
and for trial.
550. I believe it is important to note three points from the above approach to disclosure
that had a bearing on the HIT.
550.1.First, the overall scope of disclosure ordered in relation to the Horizon
Issues was, in several respects, quite narrow. But this had the effect of
meaning that some relevant material was not surfaced through the Court-
ordered disclosure, with the result that documents sometimes only came
to light late in the day as the Claimants made more requests or POL
prepared for the HIT.
550.2.Second, it was apparent to me that the Claimants’ requests for disclosure
in relation to the Horizon system evolved over time as Mr Coyne developed
his evidence during Summer 2018 and into early 2019. I do not criticise
the Claimants for this but note that it meant that the requests for, and
subsequent disclosure of, documents relating to Horizon (such as
documents from the Peak database) came out gradually, and sometimes
close to the HIT. It also meant that these requests and the corresponding
551 POL00006442.
Page 317 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
disclosure of documents continued during Autumn 2018 when I and my
team were heavily engaged in the CIT and with preparing POL's own
evidence for the HIT. This was far from ideal and not the planned
timeframe for HIT disclosure (which as set out above was intended to be
completed in July 2018).
550.3.Third, at no stage before the HIT began did the Claimants make an
application for a further disclosure order, from which I inferred that they
were broadly content with the documents that POL was voluntarily
providing.
551. For completeness:
551.1.The Court ordered a Further Issues Trial to take place after the HIT. The
relevant Order (as amended)? required disclosure of documents for that
trial to be given on a request basis — whereby the Claimants would request
a narrow class of documents and POL would provide the documents
where it considered the request reasonable ("Stage 4 Disclosure"). This
disclosure was given on 30 August 2019.5°3
551.2.In or around September 2019, POL implemented various operational
changes in order to bring its practices and procedures into compliance with
the Common Issues Judgment. This included providing frontline Helpline
staff with information on live issues with Horizon which affected the
system. This operational change resulted in new documents about bugs
in the system being produced. By a Steering Group paper dated 26
582 WBON0001566, WBON0001596 and POL00023115.
583 POL00285761.
Page 318 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
September 2019, WBD advised that these documents must be disclosed
by POL to the Claimants “as soon as possible” .5>*
551.3.By a paper dated 10 October 2019, the Steering Group was informed that
WBD had discovered that Fujitsu had not provided a number of back-
versions of KELs that were potentially of direct relevance to the HIT. As a
result, the Court and Claimants had been notified, a review was underway,
and an audit of Fujitsu’s disclosure was being scoped.*° I deal with this
further below in Section R, at §§993-1003.
(i) The Known Error Log (Q76 to Q82)
552. By Q76 to Q82 of the Request, the Inquiry has asked me about the enquiries I
made about the “Known Error Log” database (the “KEL”), and the advice I gave
POL in relation to disclosure of the same prior to the CMC in October 2017, as
well as to the CCRC.
553. I stress at the outset that the position which I advised POL to take prior to the
Autumn of 2017 was not motivated by a concern to avoid disclosure of relevant
material or to obstruct the Claimants. Based on the information provided to me
by POL and Fujitsu during 2016 and 2017 (as explained below), I did not believe
that the KEL contained material capable of calling into question the integrity of
branch accounting data. I also understood the KEL to have several thousand
entries, or “KELs”, and that they were held in a live database. This made me
believe that providing a full copy of the KEL would be a difficult and potentially
expensive exercise, and one not justified during the pre-action phase of the
554 POLO0112568.
555 POL00286050.
Page 319 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
litigation. This is why POL resisted disclosure of the KEL when responding to the
LOC and GPOC.
554. My beliefs did not change in the Autumn of 2017 in the lead-up to the first CMC.
on 19 October 2017 when it was expected that the Court would consider making
disclosure orders. I advised POL to seek Fujitsu’s agreement that the Claimants’
IT expert could be given access to the KEL. At the time, I thought inspection
would be a more proportionate, cost-effective way of reassuring the Claimants
that the KEL was immaterial than acceding to their request for a full copy of the
database. Based on what I had been told to that point, I expected Mr Coyne to
conclude that the KEL would yield nothing of particular use or value. However,
if my understanding was wrong and Mr Coyne believed the KEL was relevant,
then allowing inspection of the KEL would flush that out early and allow the
parties to discuss how the KEL (or parts of it) might be extracted and provided to
the Claimants. I thought this a pragmatic way forward. Much later it became clear
that the KEL did, in fact, contain important information, but I do not believe that
my approach prior to October 2017 was unreasonable based on what I
understood and had been told at the time.
Enquiries and advice prior to the Generic Defence
555. With reference to Q76 to Q78 of the Request, I deal first with the period prior to
service of POL’s Generic Defence.
556. In their LOC dated 28 April 2016, the Claimants referred to a “known error log”
which they understood was maintained by Fujitsu. They further believed that
Page 320 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
reports relating to it were sent to POL. To the best of my knowledge, this was
the first time that I had come across the term ‘KEL’ or ‘Known Error Log’.
557. In the course of preparing the LOR, WBD made enquiries of POL about the KEL.
Tom Porter (WBD) was tasked with gathering information from POL to help us
decide how to respond to the (32) requests for pre-action protocol disclosure
made by the Claimants in their LOC. He sent Andy Garner and Mark Underwood
of POL an email on 4 July 2016, asking whether POL “kn[ew] of a ‘known error
log’ and if so are you able to provide us with a copy?”*°"
558. Unfortunately, despite a series of chasers, we did not receive a response until 27
July 2016 (the day before the deadline for POL’s LOR). Andy Garner instructed
us that:
“With regards to the software upgrades/architecture documents and
error logs requirements, these may be with Atos however most likely
Fujitsu_being their internal records/MI [i.e. Management Information)’.
559. Since POL held little or no information internally about the KEL (and did not seem
to receive reports relating to it, contrary to what the Claimants believed),
thereafter we made enquiries directly of Fujitsu. On 19 September 2016 I sent
the following enquiry to Fujitsu:
“The solicitors for the postmasters have asked us to provide them with a
copy of the ‘known error logs' kept by Fujitsu and all correspondence
between Fujitsu and POL relating to the same. So we can respond to
this request it would be appreciated if you could confirm whether a
Horizon ‘known error log’ or a similar documents exist. We've not
decided yet on whether to provide this information (and so don't need
any documents from you at this stage) — we're just trying to scope out
what might be covered by the request. We would appreciate if you could
help with the following:
556 POL00241140.
587 POL00408698.
Page 321 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
1. What format do these logs take and where are they held?
2. What level of detail is included within the logs (i.e. are they generic
logs for all Horizon errors or linked to specific branches, do the logs
explain the consequences of the error and the fix which was applied,
etc)?
3. When did Fujitsu begin to maintain these logs and are they still
maintained?
4. Are the logs capable of being extracted and provided to us?
5. Have Post Office previously been provided with a copy of these and,
if so, to whom and when?
6. Has there been any correspondence between Post Office and Fujitsu
concerning these logs? I suspect loads — but it would be good to just
understand in high level terms how they are used.”°56
560. Pete Newsome responded on 21 September 2016 as follows:
“We would ... like to first point out our concern regarding the underlying,
erroneous, theme that these questions ... seem to be driving at. At the
risk of sounding like a “broken record”, the key premise of the HNG-X
(and Horizon) system is the Core Audit Log. This comprises the only
source of the “truth”. And to our knowledge there has been no identified
issues with the Core Audit Log and there are no KELs in respect of this
log. To this end, the questions regarding the existence or otherwise of
issues with other elements of the system are, in our opinion, a distraction
to the key premise and could be used to create an erroneous view of
HNG-X/ Horizon. ...
1. What format do these logs take and where are they held? The logs
(referred to as the Known Error Log or “KEL”) are held on a server and
contain information advising to accessor of the error condition, priority,
resolver group to handle and process to follow. Access to the log is
controlled by the accessor having to logon to the server. KELs created
as a result of an issue arising from an enterprise management event or
the result of a post incident action. KEL’s are often fixed as part of a
maintenance release and then closed. However, in some instances
KELs are not closed and remain on the system on the basis it is easier
to follow the tactical workaround procedure defined in the individual
record.
558 WBON0000460.
Page 322 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
2. What level of detail is included within the logs (i.e. are they
generic logs for all Horizon errors or linked to specific branches,
do the logs explain the consequences of the error and the fix which
was applied, etc.)? KELs are individual articles associated to individual
events. They explain the issue and any workaround for to be followed if
the issue is reported. They are logged against a priority/impact and
which capability the call should be passed onto in our service
management tool to record and resolve at the time.
3. When did Fujitsu begin to maintain these logs and are they still
maintained? The log was implemented from day 1 of the service and is
reviewed periodically. The KEL solution is currently being reviewed as
part of a service improvement activity with service managers talking with
product owners on relevance and quality of all KELs in the system,
4. Are the logs capable of being extracted and provided to us? Yes
via excel data extract
5. Have POL previously been provided with a copy of these and, if
so, to whom and when? We don’t believe POL has ever asked for this
before but it is available if required.
6. Has there been any correspondence between POL and Fujitsu
concerning these logs? I suspect loads —- but it would be good to
just understand in high level terms how they are used. To our
recollection, this has not been raised in any service meetings and as
such we are not aware of any correspondence regarding the KELs
themselves though clearly there will be numerous correspondence
regarding any issues that may have given rise to a KEL’*°° (emphasis
added).
561. There are three points to make in relation to this email. First, the message we
repeatedly and unequivocally received from Fujitsu was that if there were no
bugs or errors affecting the operation of the ‘core audit process’ (being the
process by which transaction data flows from branch to the Audit Store containing
the master record of transaction data), then there could be no possible basis on
which to impugn the integrity of the accounting data held on the Horizon system.
Second, therefore, the fact that there were no entries in the KEL relating to the
559 WBON0000461.
Page 323 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
core audit process meant that it could not logically be relevant to the intimated
claim (which, although unparticularised at that stage, could only be based on a
contention that this process was affected by bugs). Third, my main aim at this
stage was to understand the potential relevance of the KEL with a view to
determining whether it met the threshold for early disclosure in accordance with
the Practice Direction on Pre-action Conduct. I was not at that stage focused on
questions such as control of the KEL (which Pete Newsome was not qualified to
determine in any event), or the form in which disclosure might be given to the
Claimants. My main takeaways from Pete Newsome’s email were that (i) the KEL
appeared very unlikely to have any bearing on the Claimants’ claims, and (ii) POL
did not itself have copies of (or extracts from) the KEL.
562. Out of an abundance of caution so as to be clear on the first point, I responded
to Pete Newsome the same day, asking “are you 100% sure that there are no
KELs in respect of the Core Audit Log? If so, we may say this to Freeths so to try
to avoid having to disclose the KELs”.*©° He emailed two days later to say that
Fujitsu had “checked all the KELs (takes time as they have free text included)
and have found there are no KELs that directly affect the normal operation of the
Core Audit Process” .*'
563. WBD therefore informed Freeths that POL was not going to disclose the KEL at
this stage since the claim “concern[ed] errors with the Core Audit Log [and]
Following a review of the KEL FJ have confirmed that there have been no logs
in respect of Core Audit Log”.5°?
580 WWBONO000462.
561 WBON0000463.
562 WBON0001062.
Page 324 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
564. From my emails, on 31 October 2016 I was copied into a request which Paul
Loraine made for a sample of five or so KEL entries from Fujitsu.°° I believe this
was in connection with a request by the CCRC for information about the KEL
which Rodric Williams was then dealing with; since WBD already had lines of
communication with Fujitsu about the KEL because of the group action, Paul
Loraine was tasked with requesting a set of KELs for the CCRC to review.>® I
was not involved in carrying out this task and have no recollection of it, but based
on my emails I was aware that the sample KELs were provided.5® For
completeness, in January 2017 I was copied into an email by Paul to Rodric
Williams in which he had formatted various follow-up questions posed by the
CCRC into a draft email to be sent to Fujitsu.5° POL00025358, which
incorporates this email chain, shows that I suggested that Paul should lead on
gathering the necessary information from Fujitsu to answer the CCRC’s queries.
I was later copied into further emails indicating that there had been delays in
Rodric Williams sending (or authorising Paul to send) those queries to Fujitsu,”
and I believe they were ultimately ‘parked’ until after the CCRC had considered
the Generic Defence.
565. Back in November 2016, in a different context, the question of whether POL had
control of documents held by Fujitsu arose. As I have mentioned above (§399.4),
on 15 November 2016 I received an Email from Elisa Lukas (a solicitor in my
team) informing me that Rodric Williams had not sent a litigation hold notice to
563 WBONO0001069.
564 POL00249030.
585 WWBONO001069.
56 WWBONO0000473.
587 WBON0000476; POL00024817.
568 WBONO0001187.
Page 325 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Fujitsu (despite us earlier advising him to do so) because “he does not consider
their documents to be in his possession or control and it will be costly to PO”.
Elisa observed (correctly) that a document will be within a party’s control if they
have a right to obtain it, she went on to say that POL appeared to be entitled to
documents from Fujitsu albeit at a cost. I thought there was something in this
albeit the position was likely to be a little more nuanced, so I asked her to “back
this up with specific analysis i.e. what are those exact contractual rights.” Since
we had the current POL-Fujitsu contract in our possession (this having been
disclosed to the Claimants in October 2016), I asked her to arrange for the
relevant provisions to be analysed. I return to this point below, at §568.
566. On 17 March 2017, Freeths again raised the subject of the KEL in
correspondence. They contended that the draft GPOC (sent on 1 December
2016) did not confine the Claimants to an allegation that there were bugs and
errors affecting the core audit process. On that basis Freeths reiterated their
request for disclosure of the KEL, but did not explain what the Claimants’
allegations were or why this meant that disclosure of the KEL was required.°”° It
struck me that disclosure of an entire database of seemingly irrelevant material
prior to any pleadings being formally served was not required under the CPR,
and that the Claimants were seeking to effectively re-order the ordinary litigation
process in seeking extensive disclosure before pleadings. If the Claimants could
explain why the KEL was relevant to some or all of their claims notwithstanding
this, then disclosure might have been given at that point. We therefore responded
on 21 March 2017, pointing out that Freeths had “not explained the need at this
569 WBONO000154.
570 WBON0001679.
Page 326 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
stage for the disclosure of this document [the KEL] or how it will assist with
finalising the GPOC. If this document is instead required to identify issues in
individual branches then these claims should be set out before disclosure is
sought.” We stressed that POL was “being prudent about incurring substantial
costs for arguably negligible benefit and [would] therefore continue to weigh
carefully any requests for information and documents” 571
567. My email records indicate that the subject of the KEL was not revisited with
Fujitsu until June 2017 when we were in the process of readying POL’s Generic
Defence. As I have mentioned above (§481.4(iii)), a meeting took place between
me, Fujitsu, Counsel, and Elisa Lukas (WBD) on 22 June 2017, to assist Counsel
in pleading certain points of the Generic Defence.5’? A number of action points
for Fujitsu arose from this meeting. One was the requirement for Fujitsu to
confirm that the KEL was “not material to ... Branch Accounting” (being a more
widely framed question than just focused on the core audit process).5’>
568. A further action point was for Fujitsu to confirm that it did not act as POL’s agent
in relation to Horizon. I believe that this was because by this time, we had in mind
the question of whether POL was entitled to access the KEL at all. At some point,
which I believe was in the period between November 2016 and June 2017,
someone in my team (probably Elisa Lukas) told me that they had not been able
to find anything in the POL-Fujitsu contract and ancillary documents that gave
POL the right to access or take copies of the KEL. I recall that I then double
checked this myself. I was aware that contracts for the supply of IT services
571 POL00247918
572 WBONO001112.
573 POL00249567.
Page 327 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
frequently require IT companies to prepare and maintain prescribed documents,
and I wanted to be sure that we understood how the contract operated. I therefore
located and reviewed the relevant provisions in the main contract and spent one
or two hours tracing them through the various defined terms and schedules of
services to check if that gave rise to any right for POL to access or take copies
of the KEL. I recall that my conclusion was that POL did not have a general right
of access to Fujitsu documents relating to Horizon. Rather, POL’s rights of access
were limited to specific documents or categories of documents associated with
specified services provided by Fujitsu. Based on my understanding at the time
(which was that the KEL comprised Fujitsu’s internal notes on workarounds for
minor technical issues which could not affect the operation of the core audit
process), I did not believe that the KEL fell within POL’s access rights. However,
if Fujitsu acted as POL’s agent in connection with the supply of the Horizon IT
software, that might afford a different basis on which it could be said to be entitled
to the KEL. At the conference on 22 June, therefore, we asked Fujitsu to consider
this point.
569. Chris Jay (Fujitsu's in-house Counsel) followed up on 30 June 2017. He
confirmed that there was no relationship of agency between POL and Fujitsu. In
relation to our other question on the KEL, he attached an internal Fujitsu email
which recorded the question as: “whether we currently (understanding this is a
‘living’ document set) have any KELs on the audit store or that are related to
errors or bugs in the system that could cause imbalance in sub postmasters
accounts”. Chris Jay gave the answer:
Page 328 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
“As of 25 of June 2017 there were no Known Errors in the Knowledge
database relating to either the Core Audit Process or issues with branch
accounts”.5”4
570. Paragraph 50(4) of the Generic Defence was drafted on the basis of that
information provided by Fujitsu. I have described the process by which it was
drafted and signed off in detail above (at §§479-482), but in summary in relation
to this paragraph:
570.1.A first draft of the Generic Defence including this paragraph was produced
by Tony Robinson QC on 4 July 2017 and sent to me for review.9’5 An
extract containing what became paragraph 50(4) was sent to Fujitsu to
review later the same day. My covering email said “/ should be grateful if
your team could review this extract and provide any comments they may
have. In particular, if they could flag and explain any points that they
believe _not to be correct that would be much appreciated” (emphasis
added).576
570.2.In that draft, the relevant paragraph read:
“It is admitted that Fujitsu maintain a “Known Error Log”. This is
not used by POL and nor is it in POL’s control. To the best of
POL’s information and belief, the Known Error Log is a knowledge
base document used by Fujitsu which explains how to deal with,
or work around, minor issues that can sometimes arise in Horizon
for which (often because of their triviality) system-wide fixes have
not been developed and implemented. It is not a record of
software coding errors or bugs for which system-wide fixes have
been developed and implemented. To the best of POL’s
knowledge and belief, there is no issues in the Known Error Log
that could affect the accuracy of a branch's accounts or the
secure transmission and storage of transaction data. [THIS PARA
574 WBON0001134.
575 WBONO000485.
576 WBON0000491; WBONO0000492.
Page 329 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
SHOULD BE CHECKED CAREFULLY BY FUJITSU AND
CORRECTED/REWORDED AS APPROPRIATE.]”
570.3. joined a call with Pete Newsome and Torstein Godeseth on 7 July 2017
to discuss the draft. My team then made some amendments to the text
based on what was discussed on that call (but there were no amendments
to the above paragraph) and Fujitsu were asked to consider that further
draft.5’’ They did not suggest any changes to the above paragraph, and
indeed in his email approving the draft Chris Jay commented, “ail fine”,
subject to a couple of minor suggestions in relation to different paragraphs
of the text.5”8 The only amendment to this paragraph in the final version of
the Generic Defence (save for minor typographical amendments) was that
the highlighted drafting comment was removed.
570.4.1 have set out above at §§481.5-481.6 the process by which the Generic
Defence was considered internally and approved by POL and I refer the
Inquiry to those paragraphs.
571. As such, on the basis of my knowledge and instructions at the time, I understood
the KEL to be a database of internal ‘know-how’ held by Fujitsu which was not in
POL’s control and which did not bear on the issues in the proceedings. It was on
this basis that paragraph 50(4) of the Generic Defence was drafted.
Further enquiries and advice prior to the first CMC on 19 October 2017
572. Following service of the Generic Defence on 18 July 2017, matters were picked
up again with the CCRC. I note that POL00041458 shows that on 26 July 2017,
Rodric Williams sent the Generic Defence to the CCRC, who responded with a
577 WBONO001147.
578 WBON0001161; POL00249903.
Page 330 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
further set of questions relating to the KEL (posed by Grant Thornton, the
CCRC'’s instructed expert). Paul Loraine took carriage of POL’s response, which
I then commented on.5”° I set out below the CCRC’s questions in black, Paul’s
suggested answers in red, and my comments on them in yellow highlight:
“Paragraph 50(4) of the defence refers to the KEL as a “document” in
the singular, am I right in thinking it exists as a single document rather
than multiple different logs? YES IT IS A SINGLE DOCUMENT [IS
CORREC L DATABASE RATHER
SN
In terms of scale, in your email of 13/11/2016 you referred to “thousands
of entries” and “voluminous entries” in the KEL, is it possible to say how
many pages the document runs to? ANDY - PRESUMABLY ONLY FJ
COULD ANSWER THIS? DOES THIS QUERY HAVE TO GO
THROUGH THE USUAL PROCESS OF BEING PUT IN A QUEUE OF
QUESTIONS FOR FJ OR ARE YOU HAPPY FOR ME TO ASK PETE
NEWSOME (MY PREVIOUS FJ CONTACT ON THE KEL DIRECT?
ASK PETE (BUT NEED TO CC CHRIS JAY, MIKE AND
RODRI
“In your email of 13/11/2016 you stated: “we have asked Fujitsu for some
random examples of these entries, and will provide these to you in due
course”. I don’t think we have received any examples to date. I would be
grateful if we could see some sample pages, as this would give us a
clearer idea of the kind of information the Kel contains, and whether
further analysis of the document is likely to be of any relevance to our
review. WE HAVE THE EXAMPLES BELOW FROM FJ. I WILL
SUGGEST ROD PROVIDES THESE WITH HIS NEXT UPDATE EMAIL
TO THE CCRC. AGREED.
573. These queries were passed onto Pete Newsome of Fujitsu who responded by
email on 17 August 2017 as follows (Pete Newsome’s answers in red):58°
“Is it more accurate to call the KEL a database rather than a document?
Yes it is a database which changes with new KELs being added,
changed, and removed on a regular basis.
579 WBONO0000503
580 WBONO0000504
Page 331 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Is it right to say that the KEL exists as one single database/document
rather than multiple different logs? It is one database.
I understand the KEL runs to thousands of entries. Is it possible to say
how many pages the KEL would run to? Stats below. The KEL DB
Knowledge Overview lists the following i.e. 3973 KEL’s”.
574. On 18 August 2017, Paul Loraine forwarded Pete Newsome’s answers to Rodric
Williams for them to be forwarded on to the CCRC (with me in copy).5*
575. The Claimants continued to raise the matter of the KEL in correspondence (and
by way of a CPR 18 request) in July and August 2017. At around this time we
were turning our minds to the first CMC. It was appropriate at this stage to start
to consider what disclosure orders ought to be made at that hearing (albeit that
the pleadings had not yet closed as the Claimants’ Reply was not yet due), and
I have outlined the general advice that WBD gave in relation to what disclosure
orders to propose above at §§523 ff, especially §§536-537. I have also identified
the reasons why I formed the view around this time that POL should attempt to
facilitate access to the KEL (above, §554). In short, it was apparent that the
matter was not going to go away and I thought at the time that it would be
preferable to seek to put it to bed by affording access if possible. There was a
real question, however, as to whether disclosure could be managed in a sensible
and cost-effective way, given what I understood about the nature, size, and
technical content of the KEL database.
576. It was against this background that on 1 September 2017 WBD wrote to the
Claimants about the KEL in the following terms:
“Access to the Known Error Log (KEL) can also be considered as part of
these wider disclosure issues. The KEL is a not a document, but a live
and proprietary database with approximately 4,000 entries. Since the
581 WBONO001188.
Page 332 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
KEL is a constantly rolling document, the current version in use has
evolved over time and may not reflect the version in place at time which
is relevant to the Claimants' claims. Providing "disclosure" of it is
therefore not easy to do and prone to being a disproportionately
expensive exercise if not handled carefully.’**2
577. For the avoidance of doubt, by my statement that “The KEL is not a document’,
I was referring to the non-legal understanding of the word ‘document’, as a way
of expressing the practical difficulties associated with producing the KEL in the
same way that one would (for example) an email or a hard copy document. I was
not suggesting that the KEL was not disclosable — this is precisely why we made
the proposal to grant “access” to the KEL. As to the “wider disclosure issues” to
which this passage makes reference, we repeated the concern that we had
already raised in correspondence about the scope of disclosure sought by the
Claimants, having regard to both the generality of the Claimants’ Amended
GPOC and the fact that they had not yet valued their claims, which made it very
difficult to assess the proportionality of their requests.
578. On 14 September 2017, Tony Robinson QC emailed me asking to see a copy of
our 1 September letter, stating: “/ would like to understand why we seem to have
claimed that the KEL is not a document’. He suggested that it was. In relation to
this query, Amy Prime identified (rightly) the point made in the last paragraph.5*
579. Tony Robinson QC also asked why WBD had not made the point that the KEL
was notin POL’s control in our letter. As to this, he felt that we needed to be clear
that POL did not have a right to inspect or take copies of the KEL either under
the POL-Fujitsu contract or the law of agency, or as a matter of established
582 WWBONO001196.
583 WBONO0001195.
Page 333 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
practice.5*4 In relation to this second query, there followed some debate about
whether there might be some analysis by which POL could be said to have
control of the KEL. As I understood it, Tony Robinson QC’s query was concerned
with discussing within the legal team — and satisfying himself — that POL did not
in fact have control of the KEL. As noted below, I believe he was ultimately so
satisfied, because he maintained at the CMC on 19 October 2017 that POL did
not have a right to the KEL. This was also my view at the time, based on my
understanding of the contractual arrangements and the fact that POL had never
had access to the KEL (historically it had not had any access to the KEL, and
more recently it had only been provided with the odd example entry in order to
answer its and the CCRC’s queries about the format of the database). But in any
event, in practical terms I thought this point would assume less significance from
this point as long as Fujitsu agreed to allow the Claimants access. Accordingly, I
wrote in response to Tony Robinson QC that, “in principle we are happy to give
[the Claimants] access to the KEL, it’s just a question of when and how — as we
have always said to them”.5®
580. WBD then wrote on 15 September 2017 to Freeths advising them that POL were
“currently discussing with Fujitsu how to enable access to the Known Error Log.
We are hopeful that we shall be in a position to provide a constructive proposal
which would be acceptable to the Claimants by Wednesday, 20 September
2017”.586
584 WBONO0000505.
585 WWBONO001197.
586 WBON0001199.
Page 334 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
581. On 18 September 2017, I emailed Fujitsu’s legal team explaining that the
Claimants “are pressing hard for access to the KEL and are threatening to get a
Court Order. We like to discuss whether, and if so how, we could give Freeths’ IT
expert access to the KEL”. I spoke on the phone about this with Pete Newsome
on 19 September 2017 and followed this up by email dated 20 September 2017
with actions (i) for him to “speak to FJ support team about the practicalities of
doing this”, and (ii) for WBD to “circulate draft letter to Freeths with proposal for
accessing the KEL” (which I did the following day).5°”
582. On 21 September 2017 we received a letter from Freeths which stated that, as
we had not made a proposal concerning the KEL by our self-imposed deadline
of 20 September, they were now drafting an application under CPR 31.14 (i.e.
on the basis that the KEL was referred to in POL’s Generic Defence). I chased
Fujitsu with an urgent request for them to approve the draft letter to the Claimants
offering their expert access to the KEL.5**
583. Pete Newsome responded on 21 September 2017 as follows:
“Fujitsu suggests that once Freeth’s have appointed a suitable IT expert
who has signed NDA’s to preserve Fujitsu’s commercial position we
could make the system available in our Bracknell offices for supervised
inspection. Any questions can be answered on the day or submitted in
written form after the visit. Depending on the length of visit and the follow
up questions this could be a chargeable activity if not contained. ’>®°
584. By letter to Freeths dated 22 September 2017, WBD explained that, whilst we
did not agree that the Claimants were entitled to inspect the KEL under CPR
31.14 and we did not believe the KEL to be relevant, POL had discussed with
587 POL00041483.
588 POL00041483.
588 WBONO0000506.
Page 335 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Fujitsu the methods by which access could be provided to Mr Coyne at Fujitsu’s
premises.5% We asked for Mr Coyne’s availability in the next two weeks to attend
Fujitsu's Bracknell site. The letter also stated that Fujitsu had asked for Mr Coyne
to sign a routine non-disclosure agreement (“NDA”) due to the confidential or
commercially sensitive nature of the information contained in the KEL.
585. A draft NDA was prepared which provided that: “Freeths has requested access
by yourself ...of the Fujitsu proprietary technology utilised in the supply of the
services [...] known as the “Known Error Log” and of any Peak entries referenced
in the Known Error Log”. The fact that Fujitsu were insisting that the Claimants’
expert sign an NDA before he could access the KEL reinforced, to my mind, what
I had already understood to be the case, namely that the KEL was not in POL’s
control because it was a proprietary database of Fujitsu’s internal know-how on
how to manage IT systems. This was also consistent with the concern (shared
by me and Rodric Williams) that, if an order for disclosure of the KEL was made
directly against POL, its ability to comply with that order would be dependent on
Fujitsu’s willingness and ability to extract the KEL. (As I recall, this is why the
Order dated 27 October 2017 following the first CMC was made on a “reasonable
endeavours” basis.)
586. By letter dated 27 September 2017, Freeths agreed in principle to WBD’s
proposal for Mr Coyne to inspect the KEL.°%"
580 WWBON0001200; WBONO0001201.
591 WBONO0001203.
Page 336 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
587. Over the next two weeks, the parties liaised over logistics and Mr Coyne’s
NDA.5°? We wanted to make sure that Fujitsu were aware that the purpose of
this visit was strictly limited to inspecting the KEL. In this regard, Fujitsu continued
to have concerns about ensuring that the confidentiality of the KEL was
protected. Chris Jay wrote to Rodric Williams and me on 4 October 2017 stating:
“note that Mr Coyne is employed by IT Group Ltd. He may of course be
a self-employed contractor (or is it “worker"!?). I think that we need to
establish which. As you know, as an employee under a contract of
employment, conventionally this would mean that FJ enters into an NDA
with IT Group Ltd under which IT Group Ltd covenants that it has in place
legally enforceable confidentiality provisions given by him. If self-
employed it could be that he is contracted through a limited company or
even via an agency. These relationships may be more opaque from a an
enforceability of confidential obligations point of view-
Either way I would request that he enters into an individual confidentiality
undertaking
I understand that the scope is for Mr Coyne to review and inspect the
Known Error Log on a screen only and so as he can understand the
contents with no right to ask questions. [Pete has mentioned that he
would be entitled to take notes- is this correct?]
Disclosure — for disclosure only to Freeth’s for the sole purpose of
Freeth’s being able to understand the contents and thereby to conduct
its Group Litigation against Post Office?
IF entitled to take notes — the notes should be destroyed or handed back
to FJ on completion of the litigation?!"
588. Both the tone and substance of this email further reinforced my belief that the
KEL was not in POL’s control because it was an internal proprietary database
over which Fujitsu had a right to assert commercial confidence.
592 In the interests of cooperation, and with a view to narrowing the issues between the parties, we
also proposed to offer Mr Coyne access to 4,000 technical documents that described the
Horizon IT architecture, in addition to the KEL: WBON0001400.
593 POL00250828.
Page 337 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
589. Following up on the NDA point, on 5 October 2017 Chris Jay sent Rodric Williams
and me an individual confidentiality undertaking for Mr Coyne and a corporate
NDA to be signed by his employer (IT Group Ltd) to be forwarded to Freeths.
Chris Jay stated that he had “taken the view that both an individual and corporate
NDA is necessary [...] We would certainly wish an individual NDA be signed and
are more relaxed about the need for a corporate NDA... I suspect that both will
be required” 5%
590. On 6 October 2017 we wrote to Freeths to set out the specifics of the proposal
that had been agreed with Fujitsu.5°>
591. My fourth witness statement was prepared against this background. At paragraph
35, I referred to the “irrelevance” of the KEL and the fact that it was “not within
[POL’s] control for the reasons I have identified above. This evidence was
accurate on the instructions and information that we had received at that time.
This was further underscored by Counsel's sign-off on the statement. In addition,
and as noted above, Leading Counsel submitted at the CMC that the KEL was
not relevant and was not within POL’s control, based on the same information
from Fujitsu that Counsel and I had received.
592. Based on these same instructions, which I have outlined above, I went on to
describe the KEL in the following terms at paragraphs 38 to 40:
“L understand from Fujitsu that the Known Error Log cannot be easily
downloaded as it comprises data that is stored on a database, rather
than being a document in a conventional form. Unless one has the
necessary database software, reading the data in the Known Error Log
is very difficult. The alternative is to manually copy or print each entry,
but this would produce poorly formatted material and would take
5°4 POL00250841.
595 WBON0001213.
Page 338 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
significant time and work. Fujitsu believe that the best solution is for a
person with appropriate expertise to read the Known Error Log on a
screen at its offices where the information can be presented in a user-
friendly format.
To avoid incurring needless time and costs arguing about this, Post
Office wrote to Freeths on 22 September 2017 offering to arrange in the
first instance for an opportunity for the Claimants’ IT expert to inspect the
Known Error Log at Fujitsu’s premises. This offer was subject to Fujitsu’s
requirement that the Claimants’ IT expert signs a standard form Non-
Disclosure Agreement in order to project Fujitsu’s commercially sensitive
know-how that might be revealed in the Known Error Log.
However, if having seen the Known Error Log the expert believes that
disclosure of some sort is needed, the inspection process offered by
Post Office should enable Freeths to indicate precisely what is needed
and to explain why, as my firm proposed in its letter dated 29 September
2017”.
593. This evidence was accurate on the information we had at that time. Further, in
my view, the proposal we had made for the Claimants to inspect the KEL (but not
have it extracted and provided to them) was reasonable and proportionate based
on what we understood at the time — particularly given that no trial in relation to
Horizon had even been ordered at this stage, the focus being on shaping the first
trial, of the Common Issues.
594. At a Steering Group meeting on 16 October 2017, just prior to the CMC, I
provided an update on the litigation, including the following update in relation to
the KEL: “[The Claimants] are insisting on full disclosure of the Known Error
catalogue despite our offer to allow to inspect it at Fujitsu. We will be pushing
back on this point and this can be expected to be the main point of discussion at
the CMC”. I regarded our decision to push back on this point was reasonable and
proportionate on the basis of the circumstances as we then knew them and the
factors which I have outlined above.
Page 339 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
595. As noted above (§318), at the first CMC the Court made an order that aligned
with our proposal, namely that Mr Coyne’s access to the KEL should be facilitated
by him visiting Fujitsu's Bracknell site (subject to providing an NDA), leaving it
open that this could give rise to later requests for production of specific entries
from the KEL.
Disclosure to the CCRC
596. With reference to Q79.1, as I have mentioned above at §564, the CCRC asked
various questions about the KEL in late 2016 which were, so far as I am aware
having reviewed my email records, ‘parked’ in early 2017 pending POL’s Generic
Defence. I do not know whether this was agreed with the CCRC or not. Once the
Generic Defence was served, paragraph 50(4) was drawn to the CCRC’s
attention by Rodric Williams on 27 July 2017 (POL00041458, cf. §§572-574
above). Although I was copied into the emails to which I have referred, I do not
believe I advised Rodric Williams on the decision either to ‘park’ the CCRC’s
queries or to draw paragraph 50(4) to the CCRC’s attention. I was copied in
because these emails related to matters which were live in the group litigation
and other members of WBD were assisting (i.e. by gathering information for the
CCRC about the KEL), and latterly because they involved the provision of POL's
core pleading to the CCRC. As set out above at §572, when the CCRC asked
follow-up questions about the KEL based on the Generic Defence, I provided
Paul Loraine with one or two points of clarification and a steer on how to get the
other information he needed.
597. With reference to Q79.2, I did not have any other meaningful involvement in the
disclosure of the KEL to the CCRC. I continued to be copied into emails from
time to time which related to the CCRC’s questions about the KEL, and
Page 340 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
subsequently into emails which related to arrangements which were being made
for the CCRC to inspect the KEL at Fujitsu’s Bracknell Office. I cannot see that I
responded to any of these emails (which as above, I believe I was sent largely
for information purposes), save for one on 13 March 2018 which related to the
CCRC’s planned visit to the Bracknell office the following day.®® In this email, I
simply flagged to Paul Loraine that Fujitsu were currently working on a way of
extracting all the KELs (something which was not thought to be achievable in a
cost-effective way at the time of the October 2017 CMC). Following the 14 March
2018 visit POL sent Paul Loraine an update email (to which I was copied) setting
out the progress made by the CCRC so far, which referred to the visit and
commented that the CCRC had quickly established that the KEL was unlikely to
contain information which could easily be linked to a particular branch.59”
598. For the avoidance of doubt, so far as I am aware WBD's only involvement in
these matters consisted of compiling answers to the CCRC’s questions about the
KEL, and subsequently assisting in making arrangements for the CCRC’s visit to
Fujitsu. This was because we had been conducting our own inquiries about the
KEL for the purposes of the group litigation and therefore had relevant lines of
communication with Fujitsu on this issue. As a result, I understand that we were
much better placed to assist POL to obtain and prepare answers than Cartwright
King, who were generally responsible for gathering material for the CCRC and
handled all of the aspects of advice and strategy in relation to POL prosecutions.
WBD did not advise POL on any strategy for handling the CCRC, either in relation
to the KEL or more broadly.
586 WWBON0000523
597 WBON0001687.
Page 341 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(iii) The Peak Database (Q81.2, Q83 to Q87)
599. Q81.2 and Q83 to Q87 of the Request ask me about my involvement in the
disclosure of the Peak system, in particular (i) when I became aware of the Peak
system and what steps I took to investigate its content (Q84, Q85); (ii) what
advice I gave POL about disclosing that system to the Claimants, and why my
fourth witness statement did not make reference to it (Q83, Q81.2); (iii) the
background to the disclosure of some 220,000 ‘Peaks’ to the Claimants in
September 2018 (Q86); and (iv) my involvement in the disclosure of Peaks
thereafter.
Early knowledge of the Peak system
600. Broadly, my understanding (being from five years ago during the HIT) is that a
“Peak” is the term used to describe Fujitsu's internal record of an investigation
into, and (where applicable) the development of a fix for, an issue that has arisen
in the Horizon system. Whereas a KEL records (in effect) the way in which an
issue is resolved, a Peak is the record of an investigation into an issue. A Peak
is not a comprehensive record of all material generated by an investigation, since
some material may be held elsewhere (for example, in internal Fujitsu emails).
Where a Peak is generated, it is logged in a database known as the “Peak
system’.
601. With reference to Q85.1, in the course of preparing this statement I have
identified that the first time I was sent a document containing the term ‘Peak’ was
in early April 2015, when Second Sight asked POL to comment on two
documents which discussed the possibility of using the Balancing Transaction
tool to address discrepancies which had arisen as a result of the Receipts and
Page 342 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Payments Mismatch bug in 2010 (see above, §§222-223). One of these
documents stated that it related to two ‘Peaks’.5° Based on searches conducted
of my email inbox for the term ‘Peak’, I have identified that the first time it
appeared in correspondence to me (or to which I was copied) was around the
same time, on 8 April 2015, when Pete Newsome of Fujitsu provided POL with
information about the Balancing Transaction process to help POL respond to
Second Sight’s queries, which was then forwarded to me (§226 above).5°9 I am
sure that the appearance of the term ‘Peak’ in the 2010 memo and 8 April 2015
email would have meant nothing to me at that point in time.
602. The search of my email records has also shown that, when corresponding with
Deloitte on 16 August 2016 (this would have been shortly after receipt of their
preliminary report on remote access in July of that year, see above §468), I asked
Mark Westbrook: “Have you come across something called ‘Peak Incident
Reports’? I think these are some form of reporting too! used by FJ to flag major
issues. Do you know what they are and how they’re used?” However, I do not
know what prompted this query and I cannot see that I received a response to it.
603. There may well have been other references to Peaks (which is not surprising
given the sheer scale of documents moving between WBD and POL whilst the
Horizon-related matters were ongoing) but in the context of the group litigation
the first time they were raised inter partes was by the Claimants in March 2017.
At that time, the parties were in the process of negotiating a protocol to partially
598 POL00225914.
589 POL00041040. That email said that the first steps towards a Balancing Transaction being used
would be (i) that the “fi/ssue is described on a Peak incident (the incident reporting system);
(ii) that a “[rlequirement for financial correction [is] identified by Post Office and discussed with
Sub Postmaster’; and (iii) thereafter the “Peak [is] transferred from SSC to Development team
to write required correction as a script’.
600 WBON0000459.
Page 343 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
release Second Sight from their contractual obligations of confidentiality, to
enable the Claimants to discuss their cases with Second Sight (see above at
§531 for WBD’s advice to the Steering Group in October 2016 that POL should
agree to grant Freeths access to Second Sight).®' On 17 March 2017, in a letter
to WBD, Freeths wrote that one of the topics they wished to be able to discuss
with Second Sight was the Peak System (as well as the KEL).®°? It was therefore
evident that the Claimants were aware of the Peak system by at least this time.
POL acceded to the request and the protocol ultimately entered into between the
parties on 27 July 2017 expressly provided, at clause 3.1.3, that the Claimants
and Second Sight could discuss ‘peaks’:
“Subject to paragraph 3.1.5, the topics for initial discussion with Second
Sight ... pursuant to this protocol ... shall be as follows ... Errors, bugs,
fixes, issues and ‘peaks’...”.°°
Investigations and approach to disclosure leading up to the CMC
604. In the course of preparing for the first CMC, Elisa Lukas of WBD emailed Pete
Newsome (Fujitsu) on 6 September 2017 explaining that she was preparing a
witness statement regarding the documents and data relevant to the group
litigation. In relation to Horizon, she asked: “What records are kept about bugs,
how they are kept and whether these are searchable by branch affected” © This
was in the context that the parties were still considering how to structure the
litigation (e.g. whether there should be staged trials and what they should be
about) and Freeths had proposed that early disclosure be given by reference to
broadly defined generic issues, one of which was: “bugs, errors or defects in the
61 See also: WBONO0001063.
602 WBONO001679.
603 POL00250437.
604 WBON0000178.
Page 344 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
system which were or may have been the cause of discrepancies or alleged
shortfalls’.°5 We were investigating what records Fujitsu held, and how they
were held, in order to respond to the Claimants’ proposals.
605. Pete Newsome responded to Elisa on 8 September 2017. His email included the
following:®°°
“Whenever an issue is reported by a branch to the Post Office Call
Centre (NBSC) and it is deemed to be a technical issue it is passed via
the AtoS Service desk to the Fujitsu SSC Support Team. Once the call
has been passed to the Fujitsu SSC support team it is logged in the TfS
support system against a the individual branch and updated at every
stage of investigation and to its conclusion. These records are kept for
at least 7 years and recoverable by branch.
If on investigation an issue is identified which the team member deems
to need further investigation and cannot be solved either by following the
instructions in the Knowledge base or by the experience of the Service
Staff then it is logged with the 4th line support team. If an the issue is
identified that requires a programmatic fix the it is logged in a separate
system (the Peak System) and given the generic term a ‘Peak’. If it is
deemed to need new code to fix the issue (a bug) then this is produced
and fully regression tested before release into the estate via the process
outlined below. If an issue is identified which affects a number of
branches a master call record is created to consolidate any fixes or
workarounds need to be issued.”
606. He went on to explain that any proposals for system-wide software updates
arising out of Peaks are considered by Fujitsu’s Business Impact Forum (“BIF”),
and he summarised the process followed by the BIF.°” To the best of my
knowledge, this is the first time that the “Peak system” is mentioned in any
correspondence between WBD and Fujitsu relating to the litigation.
6°5 POLO00250090.
66 WWBONO000174.
607 WBONO0000174.
Page 345 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
607. Around this time, as noted above at §§580-590, we were working on ways of
facilitating Mr Coyne’s access to the KEL, an offer to that effect having been
extended to the Claimants on 22 September 2017.°8 Having reviewed my firm’s
records for the purposes of drafting this witness statement, it appears that we
were also contemplating voluntarily offering access to the Peak system (which
the Claimants already knew about) at the same time. An early draft of that
statement, prepared by WBD on or before 4 October 2017, said (page 45):
‘In addition, Fujitsu is prepared to allow the Claimants’ IT expert to
inspect the Peak System [TO BE CONFIRMED BY Fu]. With the benefit
of this information and decisions on preliminary issues, I believe that the
parties and the Court will be much better placed to make informed
decisions about the expert evidence, if any, that may be needed in the
future" (highlighting in original).6°?
608. However, that proposal was merely under consideration as we were still
investigating what was contained in the Peak system and how access could be
enabled in practice. On 4 October 2017, Elisa Lukas sent Pete Newsome a
further list of questions about the Peak system (as well as the KEL). Pete
Newsome answered these questions on 6 October 2017 (his answers in blue text
below):61°
“Peaks
How many peaks are there in the Peaks system? Open 859
Is it possible to export information from Peaks and what format would
this be in? No this is not possible without losing all the key context
Did I note correctly that many of the Peaks have no impact on branch
accounting and can relate to entirely separate parts of Horizon? Fujitsu
use the Peak system to report actions and changes required in all of the
608 WWBONO0001200.
609 WWBON0001207; WBONO0001208.
610 WBONO000507.
Page 346 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
systems required to support the Post Office HNGX solution. This
includes not just the live system but also testing and other support
systems and include all areas of the system including infrastructure and
the request of any data extracts by Post Office. The vast majority will
have no bearing on the branch accounting system
We discussed previously the possibility of giving the Claimants' access
to inspect the Peaks System as well as the KEL. Would you be happy
with us offering this? Due to the nature of the Peak system Fujitsu
believe the best course of action is to make available the Peak system
for any references that are quoted in the KEL system. The system is not
designed for running searches or clarifying Peaks for a particular
functional purpose. It also has information within it which is both
proprietary to Fujitsu and release to a wider audience could cause
security issues”.6"1
609. Pete Newsome’s answers indicated that, as with the KEL, it would be very difficult
to extract and disclose the content of the Peak database. In relation to the
question of whether Mr Coyne could be given access to the Peak system at the
same time as the KEL, Fujitsu advised that the “best course of action” would be
to make the Peak system available for Mr Coyne to inspect any Peaks which
were referred to in entries in the KEL.
610. Accordingly, when on 6 October 2017 WBD sent Freeths the details of the
arrangements agreed with Fujitsu for Mr Coyne to attend the Bracknell site to
inspect the KEL, the draft NDA provided for Mr Coyne to sign stated as follows:
"In connection with the Post Office Group Litigation, Freeths has
requested the inspection by Mr Jason Coyne [an employee
of][independent contractor to] [delete appropriately] IT Group Ltd (herein
[‘Employee"][‘Independent Contractor’] [delete appropriately] of
aspects of the Fujitsu proprietary technology utilised in the supply of the
services under the Contract and known as the “Known Error Log” and of
611 WBONO001684.
Page 347 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
any Peak entries referenced_in_the Known Error_log (collectively
“KEL”)” (emphasis added).®'2
611. With reference to Q81.2, the Peak system was not ultimately referred to in my
fourth witness statement for the following reasons:
611.1. First, the purpose of Parsons 4 was to set the context for the Court
(especially because this was the first hearing before the newly appointed
Managing Judge) to make the first set of trial directions in the proceedings,
including major strategic questions around whether there would be a
staged trial process, the directions timetable for those trials and then
disclosure in relation thereto. One part of that statement was to provide
some context around the scale and nature of the documents that might be
involved in the litigation. It sought to explain (i) the disagreements that had
arisen between the Claimants and POL regarding orders for early
disclosure, and (ii) why POL was concerned that the early disclosure
orders sought by the Claimants would require huge effort, disproportionate
costs, and be likely to produce vast numbers of irrelevant documents, in
requiring disclosure by reference to generic issues of very broad scope
(cf. paragraph 4). Its purpose was expressly not to identify the locations of
all potentially relevant documents. Thus, paragraph 57 said: “/n order to
give a sense of the difficulties involved [in identifying and recovering
relevant documents], in the following paragraphs I describe some of the
principal IT systems that Post Office has in place_and which might be
612 WBON0001213; POL00250836. Freeths returned a draft of the NDA on 13 October 2017 with
this section substantially unchanged. This wording was in substantially the same terms in the
final agreed NDA.
Page 348 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
relevant _to_this litigation" (emphasis added). This was not intended or
expressed to be an exhaustive list.
611.2.Second, and as I have mentioned, the Claimants were well aware of the
Peak system and had not made a request for disclosure or inspection of it
(but were permitted to discuss it with Second Sight as a result of the
protocol the parties had put in place).
611.3. Third, there was no need to refer separately refer to the Peak system
when setting out POL's offer to facilitate Mr Coyne’s access to the KEL,
since it was inherently part of that offer and not a separate inspection.®"?
612. The approach we adopted in relation to the Peak system was therefore:
612.1.To arrange with Fujitsu that Mr Coyne would be able to access the Peak
system to inspect any Peak he identified from the KEL, and to include the
Peak system in our EDQ which was served on 6 December 2017. This
recorded that: "If Fujitsu identifies an issue in Horizon that requires a
programmatic fix then it is logged in its database, the Peak System, and
labelled as a ‘Peak"".5'4
612.2.Otherwise, to leave it to Mr Coyne to inspect individual Peaks as he saw
fit during his visit to Fujitsu in November 2017, and/or to the Claimants to
seek further access to, or disclosure from, the Peak system if they wished.
613 On 7 October 2017, when commenting on the statement Tony Robinson QC queried offering
the Claimants inspection of the whole Peak system. As set out above, the day before we had
offered a narrower form of access to the Peak system as recommended by Fujitsu (namely,
access to Peaks Mr Coyne wanted to inspect which he identified from KEL entries). I therefore
advised Tony Robinson QC that the proposal had now been dropped and was a redundant
hangover from an earlier version of the statement, although I expected the Claimants to seek
inspection along these lines in due course: WBON0000190.
614 POL00252048.
Page 349 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
613. This was, in my view, a reasonable approach based on the information and
understanding I had about Peaks at the time (as set out above).
Background to September 2018 disclosure
614. In the event, and notwithstanding that Mr Coyne could have requested access to
any entries in the Peak system he identified from the KEL during his November
2017 visit, the Claimants did not raise again the issue of the Peak system until
11 April 2018, when it was discussed at a joint meeting between the parties and
their respective IT experts (referred to above at §544).°' The purpose of that
meeting was to agree (i) the scope of any further information or documents
relating to Horizon that the experts required, and (ii) a process for the experts to
inspect the Horizon system. The Peak database was identified by the experts (as
was the KEL) as a potentially useful source of information from which bugs could
be identified, although the parties had different ideas about how that information
should be mined. At the risk of oversimplification, the Claimants appeared to
prefer a top-down approach involving searching for key words and flags
associated with bugs generally, in order to capture potentially relevant
information which could then be examined; POL believed that the Claimants
ought to identify suspected problems they had encountered with Horizon (e.g.
anomalous transactions, unexplained behaviours of the system, etc.) which
could then be the focus of more targeted enquiry to see if those problems were
caused by bugs. I was surprised to learn at this meeting that Mr Coyne did not
appear to have been provided by the Claimants with any specific details of their
allegations, or examples of problems they had encountered.
615 WBON0001235.
Page 350 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
615. Following that meeting, Freeths requested access to the Peak system and I
recommended that POL should seek to agree this with Fujitsu.°"* POL agreed
and by a letter to Freeths dated 17 April 2018, confirmed that it would facilitate
access to the Peak system if Mr Coyne wished to inspect it.6'7 No response was
received to that offer and WBD chased Freeths on 14 May 2018.8 They
responded on 18 May 2018, confirming that Mr Coyne sought a session at Fujitsu
to inspect the Peak system as well as a separate system known as the TFS.®'9
The draft Order attached to Freeths’ letter contemplated that Mr Coyne’s
inspection would take place over two days
616. On 21 May 2018, Jonny Gribben (who was leading on work relating to the
preparation of the experts’ reports) emailed Pete Newsome of Fujitsu with me in
copy, seeking to make arrangements for the experts to attend Bracknell to
inspect the Peak system and the TFS. I note that Jonny asked Pete Newsome
whether both systems could be inspected “in a day’, to which Pete Newsome
responded yes.®2°
617. On 7 June 2018, Freeths informed us that Mr Coyne was available on 14 and 15
June 2018 for the inspection.*@' However, Dr Worden was not available then,
but was available on 21 June 2018 or any day during the week commencing 25
June 2018.52 Jonny therefore replied to Freeths on 12 June 2018 offering these
616 POL00022706; cf. §§544-545 above.
617 POL00254578.
618 POL00254961
619 POL00254995; POL00254996
620 WBON0001247
£21 WBON0000529.
622 WBON0000530
Page 351 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
dates instead.®**- Freeths then responded to say that Mr Coyne was available
on 27, 28 or 29 June 2018.54
618. With this Jonny enquired after Fujitsu's availability to accommodate an
inspection, and Pete Newsome advised that 28 June 2018 was feasible and
reiterated that he believed the Peaks and KELs could both be inspected in one
day.®2°- Jonny informed Freeths and they replied saying that Mr Coyne thought
that two days were necessary, so would want to return for a second day.
However, they said 28 June 2018 was in his diary and the inspection went ahead
on that day.®26-
619. In the meantime, at the CMC on 5 June 2018, Mr Justice Fraser had ordered that
the experts be given two days for inspection.®’ Ultimately, on 26 June 2018 when
the experts formally submitted their joint Requests for Information (“RFI”)
pursuant to the CMC Order, Mr Coyne’s request for inspection was in the
following terms: “[ojne day agreed at present, second day to be arranged”
(Request 1.9).628 So far as I am aware based on searches carried out of my email
records, Mr Coyne did not in fact request a second session to be arranged after
this. I also note that an email sent to me and Jonny Gribben the day after the
inspection by Chris Emery (Dr Worden’s assistant) recorded Chris’ view that
“[o]verall yesterday went well. We achieved all of our objectives and learned a
623 WBON0000531.
624 WBON0000532.
626 WBON0000533.
626 WBON0000534
£27 POL00120352.
628 POLO0110998
Page 352 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
great deal about Peak and TfS. We can consider those systems well and truly
inspected!” ©
620. With reference to Q86.1, I therefore believe that that the account in Freeths’ letter
of 2 October 2018 (POL00003386), which characterised POL as only being
prepared to agree a one-day visit, is not a full and fair description of the approach
taken by POL or WBD to facilitating access to the Peak database. The reality of
the situation, so far as I can tell based on the records available to me, was that
an initial day was pragmatically agreed with the experts and Fujitsu, and in the
end Mr Coyne did not require a second day in which to inspect the Peak system.
621. Following the inspection on 28 June 2018, on 20 July 2018 Mr Coyne submitted
requests for disclosure of a selection of records from the Peak and TFS systems,
which included requests for POL/Fujitsu to identify “(v) PEAK and/or TfS records
where the error or issue resulted in financial impact to either Post Office or a
Subpostmaster’, and “(vi) PEAK and/or TfS records for any Claimant who has a
record including any audit data for the period (at least a month) of the PEAK/TfS
record’.° In our view these requests were not consistent with Model C
disclosure. The first was not a class of document but required POL to self-
determine which Peaks and TFS records recorded bugs by reference to
undefined criteria. The second, whilst unclear, did not appear to correlate with
any of the Horizon Issues. Our expert, Dr Worden, who had by that time
inspected the Peak system, had similar views. In his view:®"
“Request (v) - Coyne is basically asking WBD to do his job for him. There
are_a large number of PEAK and TfS records, and the job of filtering
628 WWBONO0001256.
690 WWBONO000559.
631 WBON0001273.
Page 353 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
them all to see which ones have financial impact is what the experts
need to do - it would be disproportionate to ask FJ to do such a large
complex manual job.
I am tempted to suggest - ship Coyne a dump of the entire PEAK
database, which he can run up on SQL server or the appropriate DBMS.
Or grant him read-only access to the PEAK system, if that is possible.
Then he can search it himself. A database dump can't be that big - I bet
it fits on a disc of a few TB. (If WBD did that, we might be given the same
thing but I don't know that we'd do much with it. Rather than set up a
database, we'd prefer to access the PEAK system remotely)
Request (vi). This does not make sense - or more politely, the request is
not clear. PEAK and TfS records are not ‘for claimants’ or ‘for individual
subpostmasters'. He apparently wants all PEAKs and TfS which span
certain periods. If Coyne sends a list of months of interest to him, it might
then be possible to do an automatic filter of PEAK and TFS records
covering those months. But given there are 250,000 PEAK records for
216 months, if Coyne selects even 20 months this will give him about
20,000 PEAKs to look at. You might as well send him the lot, as
suggested above.”
622. Thereafter, during August 2018, Jonny Gribben liaised with Fujitsu to establish
what could be provided. As I understand it, these enquiries essentially confirmed
that Mr Coyne’s requests were not practically possible to comply with on their
terms, but that a summary of all of the Peak entries (which numbered
approximately 220,000) could be extracted if Fujitsu developed a bespoke
application for this purpose. These could then be searched by the experts using
their own software (as Dr Worden suggested) and more focused requests could
then be for the full entries for any Peaks which they identified as being of
interest.6°2
623. The deadline for POL to either comply with Mr Coyne’s request dated 20 July
2018, or notify the Claimants that it was objected to, was 8 August 2018 (i.e.
whilst these enquiries were ongoing): see Mr Justice Fraser’s Order of 25 July
832 WBON0001277; WBON0001278; WBON0001279; WBON0001280; WBON0001294.
Page 354 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
2018.53 Consequently, on 8 August 2018 WBD informed the Claimants that Mr
Coyne'’s fifth and sixth requests were objected to, adding:®*4
“You will see in our client's responses that it is working with Fujitsu to
establish whether a mechanism could be created to export or provide Mr
Coyne with direct access to the 220,000 Peak entries or at least some /
part of them. We believe that this may assist Mr Coyne but there are
serious technical barriers to doing this which are currently not possible
to overcome. We will revert when we are able to do so, but in the
meantime we remind you that there is an open offer for the experts to
inspect the Peak system at Fujitsu's offices for a second day”.
624. Incidentally, with reference to §620 and my answer to Q86.1 above, I note that
WBD here reminded the Claimants that it was open to Mr Coyne to spend a
second day inspecting the Peak system. To my knowledge, he did not do so.
625. Fujitsu confirmed that it could effect the extraction (and what the parameters for
doing so would be) on 20 August 2018 and Jonny asked them to arrange this as
soon as possible.®*> They were extracted and provided to WBD on 31 August
2018, after which they were ingested onto Relativity (which we were using to run
electronic disclosure exercises in the group litigation).
626. On 26 September 2018, WBD prepared the Steering Group Decision Paper
referred to above at §548, advising on the issue of whether the entire (now-
extracted) content of the Peak system should be disclosed voluntarily, i.e. in the
absence of a valid Model C request from the Claimants and/or an order for
specific disclosure.®** The briefing paper contained the following advice:
633 WWBON0001326.
634 WWBONO0001668; POL00256155
635 WWBONO0001294.
636 The meeting on 26 September 2018 being the first available opportunity to obtain a decision
from the Steering Group. There were two meetings earlier in September but there appears not
to have been time in the agenda to discuss the Peaks issue, as they were concerned with the
pressing issue of POL’s application to strike out parts of the Claimants’ witness evidence for
the CIT: see POL00256731and POL00257086.
Page 355 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
“Whilst Post Office has not been ordered by the Court to provide
disclosure of the Peak System, there are a number of reasons why Post
Office should consider providing these documents to the Claimants now:
Providing voluntary disclosure on this scale (c.220,000, documents) will
be viewed favourably by the Managing Judge and continues Post
Office’s approach of providing assistance to the Claimants where it is
reasonable and proportionate to do so.
It neutralises some of the Claimants’ Expert’s requests for information,
because the answers can be found in the Peaks.
These documents will be of assistance to Post Office’s expert who
cannot be provided with them unless they are also available to the
Claimants’ Expert.
Providing disclosure of these documents now will enable the Claimants
Expert to take these into account when producing his expert report (due
on 16 October 2018). Post Office will therefore have visibility of the
Claimants’ Expert's position at an earlier stage in the proceedings.
Post Office has an ongoing duty to disclose adverse documents. Given
the nature of the documents contained in the Peak System, it is likely
that it will contain adverse documents and therefore, disclosure of these
will be needed to be given at some stage.
LJ
It should be noted that providing disclosure of the Peak System three
weeks in advance of the Claimants’ Expert's report may trigger criticism
that this volume of documents were not disclosed at an earlier stage.
These risks can be neutralised to a certain extent through reminding
Freeths that there has been an open invitation for the Claimants’ Expert
to inspect the Peak system, an opportunity which they have not taken
up, and that Fujitsu have had to develop a unique programme to enable
documents to be extracted, which is taking time. Further, due to the lack
of crystallised allegations against Horizon we cannot determine
relevancy and therefore it is necessary to disclose all c.220,000
documents”.
627. The Steering Group accepted WBD’s advice and instructed us to disclose the
extracted Peaks, which we did the following day (27 September 2018). Our
rationale for disclosing them was as we advised in the briefing paper cited above,
Page 356 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
and the timing is explained by the background set out above and reflected in the
Steering Paper. In answer to Q86.3, the timing of disclosure was absolutely not
calculated to disrupt. The facts speak for themselves:
627.1.The Claimants were aware of the Peak system from the outset of the
litigation. Mr Coyne had the option of viewing some Peaks when he
inspected the KEL on 28 November 2017.
627.2.The Claimants did not seek a copy of the Peak system as part of the
disclosure ordered on 5 June 2018 (or at any time before that) although
they were well aware of the system and their expert was given access to
it, as set above.
627.3.The first request for disclosure relating to the Peak system was Mr Coyne's
request on 20 July 2018, and that was not a request for the whole system
but for an unidentifiable subset of peaks (and in any event the Court's
Order permitted the experts to make requests for information about the
Horizon system, not to seek disclosure of documents).
627.4.We entered into discussions with Fujitsu about extracting and disclosing
the Peak system on 1 August 2018, having concluded that Mr Coyne’s
requests of 20 July could not sensibly be complied with, and having
established that our expert was of a similar view.
627.5.In line with the Court-ordered deadline for doing so, we notified the
Claimants of POL's objections to Mr Coyne’s request on 8 August 2018.
We also informed them of the objective to extract the entire database of
220,000 Peaks, noting the uncertainties around this given the
technological barriers in place at the time. They could not therefore have
Page 357 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
been surprised when the Peaks were disclosed and nor did they ask for
that disclosure to be given by a particular date.
627.6.The Peaks were disclosed to Freeths as soon as practicable thereafter, as
explained above.
627.7.The Steering Group paper records that we believed disclosure of the Peak
entries was a helpful step, and explicitly not a litigation tactic designed to
disrupt the Claimants.
628. As to Q86.4, an integral part of any disclosure exercise is to review documents
for privilege. Each document must be reviewed manually by someone competent
to do so to identify whether it is privileged. Any privileged material within the
document must then be redacted; and those redactions are typically then
reviewed by another member of the document review team. It would have taken
months for the WBD team to manually review all 220,000 Peaks for privilege,
and we were concerned to ensure that the Claimants had access to the Peaks
as soon as was reasonably and practically possible. We therefore decided that
the most prudent approach would be to apply appropriate search terms relating
to the word “privilege” to the 220,000 Peaks and hold the responsive documents
back to manually review them. This ran the risk that certain Peaks which
contained privileged material but which did not respond to the search terms
would be disclosed, but we considered that the imperative of disclosing the
Peaks promptly overrode this concern. 3,866 documents responded to our
search terms and we therefore disclosed the remainder (approximately 218,000)
on 27 September. We conducted a manual review of the responsive documents,
as a result of which we identified that only four redactions needed to be applied.
Page 358 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
The balance of the 3,866 documents were disclosed with these redactions
applied on 25 October 2018.°°”
629. It follows that we did not assert privilege over 3,866 documents. We temporarily
held back 3,866 documents to manually review them for privilege because they
were responsive to relevant search terms. As soon as we finished that manual
review, we disclosed the 3,866 documents with four redactions applied for
genuinely privileged material. In adopting this approach we prioritised the swift
disclosure of the majority of the Peaks to the Claimants, even if this ran the risk
of disclosing documents which, on a manual review, would have been
determined to have been privileged.
Subsequent disclosure of Peaks
630. With respect to Q87 of the Request, I do not believe that there were any further
substantive tranches of Peaks disclosed during the litigation. Nor do I recall the
Claimants seeking disclosure of further tranches of Peaks (but I have not been
able to review all the inter partes correspondence in the time available to confirm
this). There were no further orders made by the Court directing disclosure of
more Peaks. I can see from my emails that my team identified that Mr Coyne
referred to seven Peaks in his expert report that he said had not been disclosed.
This triggered us to make inquiries about these Peaks with Fujitsu,®°* and we
were able to determine that these Peaks were not in the Peak system, but it was
possible to retrieve some information about them from a back-up archive. That
information was provided to the Claimants.®°° Other than the disclosure of Peak
637 WBON0001340.
68 WWBONO0001431.
638 WBON0001429.
Page 359 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
PC0273234 which I address below, I do not believe that there were any more
disclosures of one-off Peaks but I have not had time to review all the disclosure
lists to confirm this.
631. As to Peak PC0273234: on 25 February 2019, the IT experts filed their second
joint statement. Page 25 of the statement contained a reference to a “Drop and
Go Bug” (also termed Bug 28), the effect of which was recorded in Peak
PC0260269. Fujitsu provided WBD with comments on this bug in a note dated
25 March 2019:5°
“Fujitsu are not aware of the Drop & Go business process so are unable
to comment on whether this was a user error or fault with the ATOS
APADC script so call PCO260269 was routed for ATOS investigation.
Fujitsu are unaware of any further updates against this call, however,
subsequent _call_PC0237234 raised August 2018 with the same
symptoms would suggest that the issue may still remain unresolved”
(emphasis added).
632. On 29 March 2019, Katie Simmonds of WBD informed Fujitsu that WBD had
been unable to locate Peak PC0273234 on Relativity. In response, Matthew
Lenton of Fujitsu stated:*"
“It looks like PC0273234 may have been raised at around the time the
Peak extract was created in August 2018, so may not have been
included. A copy is attached.”
633. On receipt of the Peak, on 29 March 2019, Katie wrote to Atos in order to obtain
more information about the Drop and Go Bug. On the same date she also
emailed Peak PC0273234 to WBD’s Amy Prime and Charlie Temperley asking
for confirmation of whether the document had been disclosed and if not, for it to
be added to our internal disclosure tracker.*42 She notified me of having done so
40 WBON0000217.
®41 WBONO0000211.
&42 WBONO0000203.
Page 360 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
on the same day.®3 On 1 April 2019, Amy Prime emailed Charlie Temperley
asking him to add these three documents to the disclosure tracker.°44
634. The purpose of the disclosure tracker was to ensure that any potentially adverse
documents which came to light and therefore needed to be disclosed were
logged and actioned. Such documents tended to be provided to WBD in a
piecemeal way from a variety of sources. As documents were received, they
were gathered together, reviewed, and the disclosable documents were then
disclosed at intervals as a pack of documents. We considered this to be a more
manageable and reasonable approach than drip-feeding documents to Freeths.
635. Meanwhile, Katie sought an explanation from Fujitsu about why this Peak was
not identified sooner.®45 From 2 April 2019, she also began making enquiries with
Atos in order to ascertain whether this was a further incidence of the Drop and
Go bug.*46
636. Katie then made enquiries to satisfy herself that Fujitsu's explanation for the late
production of Peak PC0273234 made sense. She wrote to Charlie Temperley on
5 April 2019: “This is the explanation from FJ below RE the Peak. Does this align
with the extraction time we have?” Charlie responded on 5 April 2019: “.../ think
so yes. Sorted by document date, the last Peak on Relativity is PC0273139 dated
17/08/2018 12:45,"°47
637. Following this, my team continued work on investigating issues arising from
Peaks, related expert evidence and the (then) adjourned Horizon Issues trial.
&43 WWBONO000218.
644 WBON0000276
®45 WWBONO0000275.
&46 WWBON0001537; WBON0001535.
647 WBON0000250
Page 361 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Katie Simmonds raised specific queries with Atos in May 2019 in relation to Peak
PC0273234 and pursued those with follow up enquiries.*4® Also during the
course of May 2019, Katie was in regular contact with Counsel in relation their
ongoing work on the bugs which were the subject of the Horizon Issues Trial. My
understanding is that she would likely have discussed both Peak PC0273234
and her investigations into whether it was further evidence of the Drop and Go
Bug with Counsel.
638. On 22 May 2019, Katie emailed Counsel with seven additional “bug reports”
(being summaries in respect of the bugs in issue in the proceedings which WBD
prepared to assist Counsel).®49 One of these bug reports related to the Drop and
Go Bug.®° Within the report, WBD stated that Peak PC0273234 was being
disclosed. Katie explained that she had:
. included our queries regarding disclosure and the addition of
documents to the trial bundle in the below table ... In terms of timings for
adding documents to the trial bundle, we propose to keep a record of
any and add these to the list Tony [Robinson QC] sends through before
trial so that everything can be added in one go, unless you need the trial
bundle references sooner’.
639. The table in Katie’s email explained in relation to the Drop and Go Bug that
“There are several new documents that we are in the process of disclosing ...”.
She asked for confirmation of which documents should be added into the trial
bundle. On 22 May 2019, she also emailed Charlie Temperley asking him to
confirm that all new documents referred to in the bug reports were on the
disclosure tracker.°5
648 WBON0000225; WBON0000273; WBON0001581; WBON0001603.
&48 WWBONO001608.
50 POL00132736
51 WBON0000227.
Page 362 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
640. On 23 May 2019, Counsel (Simon Henderson) replied to Katie Simmonds’ email
of 22 May 2019 providing general guidance as to the further documents to be
disclosed in relation to the bugs:®°2
“... we should not be providing further disclosure unless we are under
an obligation to do so (adverse doc or should have been previously
disclosed etc) ... some of the docs may not be seen as helpful to PO’s
case. However, if, as seems to be the case, they are part of the story of
that bug, I think it is artificial to omit them. I think the intention is to upload
a big batch of docs: these should form part of that batch” (emphasis
added).
641. On 28 May 2019, Charlie Temperley confirmed that the documents to be
disclosed had been added to the disclosure tracker.® Also on 28 May 2019,
Charlie emailed me (copying Katie Simmonds and Jonathan Gribben), stating:*4
“In the absence of Amy and Lucy, I'm monitoring disclosure tasks this
week. After we spoke about this earlier I sorted out my disclosure tracker
(to separate out the trial 3 stuff which is not a problem for this week). The
outstanding HIT documents for disclosure this week are set out in the
table below FYI.”
642. Charlie set out a table of outstanding documents for disclosure, which included
Peak PC0273234.This email reflected the fact that I had asked Charlie to check
on the status of documents on the disclosure tracker that might need to be
disclosed. Charlie emailed me again on 29 May 2019 to confirm that he had
received another adverse document from Rodric Williams of Post Office that we
had been waiting for. Peak PC0273234 and various other documents were then
disclosed on 31 May 2019.95
852 WBON0000290.
&53 WWBONO000251
&54 WWBONO000159.
655 WBONO0000149.
Page 363 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
643. On 17 June 2019, Tony Robinson QC reviewed a draft of my eighteenth witness
statement (which Mr Justice Fraser had directed to be produced to explain, inter
alia, the late disclosure of Peak PC0273234) and raised two queries. Amy Prime
wrote to me and others in the WBD team on 17 June 2019 setting out Tony
Robinson QC’s queries (in capitals) and then her responses and her additional
queries alongside them as follows:®°>
“[CAN WE ADD A SENTENCE EXPLAINING WHY IT TOOK FROM 3
APRIL TO 31 MAY FOR THE DOCUMENT TO BE DISCLOSED, TO
FORESTALL A DEMAND FROM THE JUDGE FOR A FURTHER
WITNESS STATEMENT EXPLAINING THIS?] The Peak was provided
to Katie by Matthew on 29 March 2019 (email attached), the documents
were processed into [Relativity] on 4 April, but my understanding is that
we waited until the bug investigations were substantially complete and
we had what we believed to be all of the new documents which we
thought would come out of these investigations and then disclosure was
given all in one go. Does this match your understanding? This comes
partially from Katie's email to Charlie on 22 May 2019 (attached) which
just predates the disclosure of this peak.”
644. On 18 June 2019, Katie Simmonds wrote: “This aligns with my understanding —
I'm not aware of any further additional Peaks which have been disclosed/
provided by FJ separately.”°5”
645. With respect to Tony Robinson QC’s second question, Katie wrote:
“As per our call, this isn't something that we discussed with Counsel and
documents were simply being sent and added to the disclosure tracker
to be disclosed as and when the next list was run but appreciate this is
something that was discussed between you and Charlie in terms of
waiting for a series of adverse documents to be disclosed in one go, as
opposed to simply drip feeding these.” (emphasis added)
£56 WWBONO0000293.
657 WBON0000138.
Page 364 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
646. Amy Prime then emailed Counsel on 18 June 2019 with the following
summary:®8
“...On the timing of the disclosure, PC0273234 was disclosed due to it
being a known adverse document. WBD waited until the investigations
into the Coyne bugs were substantially completed and therefore the
likelihood of further disclosure being uncovered was thought to be low.
Once we had this batch of documents that had come to light as part of
the investigations into the bugs, disclosure of them was provided all
together. With hindsight, yes the document could have been disclosed
sooner but we were seeking to not drip-feed documents to the Cs ona
daily basis” (emphasis added).
647. With hindsight, I accept it would have been preferable if this Peak had been
disclosed more quickly:
647.1. Our approach was that (i) new potential documents to be disclosed were
assembled on the disclosure tracker, (ii) the documents were reviewed
and investigations undertaken into whether they were disclosable and (iii)
then disclosure was given of disclosable documents in batches rather than
on a drip-feed basis. The same approach was adopted in relation to this
Peak. In context, at the time, this Peak did not appear to be of particular
importance. Indeed, it is worth noting that when dealing with the Drop and
Go Bug in his Technical Appendix, the Mr Justice Fraser did not refer to
Peak PC0273234 (see paragraphs 402 to 410).
647.2. At this stage, a lot of activity was going on in parallel: we were in the
middle of the Recusal Application which had been issued on 21 March
2019 and was heard and dismissed on 9 April 2019, then resulting in
urgent attempts to appeal that decision, leading to the Court of Appeal’s
658 WBON0000139.
Page 365 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
refusal of permission to appeal in early May 2019. Alongside this, the HIT
had been adjourned on 21 March 2019, before resuming briefly on 11 April
2019 and then being adjourned to 4 June 2019.
647.3. Peak PC0273234 was not held back as part of some deliberate attempt
to gain a tactical advantage, nor was there a failure to observe professional
standards of behaviour on either my part or that of my team, who were
working extremely hard at this point. The time taken simply reflected (i) the
extreme pressure that we were all under at that stage and (ii) the general
approach that we had adopted of seeking to investigate and understand
what this Peak related to and then making disclosure in one batch rather
than on a piecemeal basis.
(iv) Disclosure of reports generated by Project Zebra (Q88.3)
648. I have dealt with my (very limited) involvement in the instruction of Deloitte in
relation to Project Zebra and my awareness of the six reports referred to in Q88
(only two of which I received prior to receipt of the Request, one in August 2014
and the other in March 2016): see above, §212 and §299. Here I deal with the
question posed at Q88.3 of the Request, namely, whether I considered that the
issues raised in the reports I read should be disclosed to (i) the Claimants in the
GLO, or (ii) convicted SPMs.
649. In sum, the reports of which I was aware (being the Project Zebra Desktop
Report, POL00028062, and the Board Summary, POL00028069) were not
disclosed to the claimants in the GLO proceedings because they were privileged.
Project Zebra was commissioned following advice from Linklaters in connection
with the risk of civil proceedings being brought against POL. Thus, the dominant
Page 366 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
purpose of the Project Zebra reports was to obtain advice in respect of litigation
which was in reasonable contemplation, and they were not therefore required to
be disclosed to the claimants in the GLO. As for whether the “issues” revealed
by these two reports were disclosed, the material issues which they raised
concerned the Balancing Transaction functionality and (the potential implications
of) privileged user access which certain Fujitsu personnel had. These findings —
which were less far reaching than later investigations revealed Fujitsu's remote
access capabilities to be — were the subject of admissions in POL’s Generic
Defence, and to this extent were ‘disclosed to the Claimants in the GLO.
650. As for whether these issues ought to have been disclosed to convicted SPMs,
this was not a matter on which I was instructed or competent to advise. I am
aware that consideration was given to these matters by Cartwright King and Brian
Altman QC in 2015, and that they were subsequently revisited by Brian Altman
QC in 2016 following advice by Jonathan Swift QC. I have set out the extent of
my involvement in those instructions elsewhere in this statement: see in
particular §§229 ff, §§296-297, and §§458 ff.
(v) Redaction of documents (Q90.1, Q91 and Q95.1)
General methodology
651. As the Partner in charge of the litigation, I approved the approach to be adopted
in respect of redactions and I was the highest point of escalation within the WBD
team for more difficult issues that could not be resolved by more junior solicitors.
This was the extent of my direct involvement with redactions. In a disclosure
exercise of this scale involving millions of documents, some 537,040 of which
Page 367 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
were disclosed, it would not have been practicable for me to have had more
involvement than this.
652. As to the approach which I approved, it was typically as follows.
652.1.Stage 1: Our paralegal team undertook a first level review for relevance
and privilege. Prior to embarking on each review exercise, the team would
attend a briefing call and be provided with a briefing note which explained
privilege.69
652.2.Stage 2: Qualified lawyers at WBD undertook a second level review of all
documents which were tagged for disclosure or tagged as relevant but
privileged, as well as a spot check of approximately 10% of documents
which had been tagged as not needing to be disclosed because they were
irrelevant. Where a document had been flagged as relevant but part-
privileged, therefore requiring redactions to be applied, these redactions
would be determined by the qualified lawyer (although the first level
reviewer may have left a comment on the document explaining where the
privileged material was located).
652.3.Stage 3: If a second level reviewer was unsure about the approach to be
taken in relation to a particular document, the document would initially be
flagged to Amy Prime for review and then, if a further review was required,
escalated to me.
652.4.Counsel: There was also an escalation process whereby Counsel was
asked to review redactions which were applied to a small number of
documents prior to their disclosure, where the application of redactions to
659 For example: WBON0000141, which annexed WBON0000145.
Page 368 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
privileged material referred to in non-privileged documents was not clear-
cut.
653. This approach was applied in most circumstances, but as disclosure was given
in tranches it may have been adapted, for example the first stage was not applied
in every instance so the first level review for some documents was undertaken
by qualified lawyers instead of paralegals. Further, we adopted a slightly different
approach in respect of disclosable documents which were to be redacted to
remove material which was irrelevant and confidential. I deal with this separately
below (at §§666-671).
654. I note that only a tiny proportion of documents were redacted (fewer than 500 out
of over 530,000 disclosed, i.e. less than 0.1%). Of those, only a fraction were
deployed in Court.
Redactions to the Project Zebra Action Summary
655. I can only specifically recall one document that was escalated to me by my team
as part of the process outlined above, in the course of my team’s preparation for
the Common Issues and Horizon Issues Trials (putting aside redactions that
Freeths or the Court asked POL to revisit —- addressed further below). However,
I do not now recall all redaction issues that might have been raised with me given
the passage of time, and I would have given my team informal guidance and
reminders about the nature of privilege and the approach to be applied
throughout.
656. The document which I recall being escalated to me was the Project Zebra Action
Summary referred to at Q91 (the “Action Summary”), of which POL00027054
is the unredacted version and POL00002356 is the redacted version. The Action
Page 369 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Summary contained references to the “Deloitte report’ (being the privileged
Project Zebra Desktop Report) and the substance of the recommendations in it.
I discussed the Project Zebra Desktop Report above at §§212 ff, and at §649, I
outlined why I considered that this report was privileged.
657. During the disclosure exercise for the CIT, my team came across the Action
Summary, which was an internal operational document that had been produced
as part of POL’s workstream to implement the recommendations contained in the
Project Zebra Desktop Report. It fell within one of the classes for Stage 3
Disclosure and was therefore prima facie disclosable, but upon reviewing it my
team felt unsure whether it was wholly or partially privileged by reason of the
repeated references to the “Deloitte report’ and that report's recommendations.
This was escalated to me and I too had my doubts, so we sought Counsel’s
opinion.
658. Tony Robinson QC responded on 26 July 2018, expressing his and Simon
Henderson's preliminary view that: “privilege c[ould not] be claimed over the
whole of the Zebra Action Summary but the parts which repeat or summarise the
contents of the Deloitte report c[ould] be redacted.”*' He identified a number of
further questions that he required answers to before he could reach a final
conclusion, as follows:
“1. Is it definitely the case that the dominant purpose for which the
Deloitte report was obtaining legal advice about or obtaining information
in connection with the conduct of anticipated litigation? [...] What we are
hoping to receive is factual instructions allowing us to be clear that the
Deloitte report was not created for two equal purposes, one of which was
anticipated litigation and the other was making Horizon more robust...
£0 WWBONO001264.
®61 WBONO0000567.
Page 370 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
2. Who produced the Zebra Action Summary, who was it produced for
and for what purpose was it produced? To whom was it disseminated
and what actions were taken as a result of it?
3. What was the source and purpose of the “recommended
remediations” set out in the Zebra Action Summary For example, were
they recommendations produced by management of changes to be
made in the systems around Horizon with a view to giving effect to
recommendations of a more general nature made in the Deloitte report?
4. Were these “recommended remediations” actually implemented?
5. Who were the parties to the email you attached (Julie George, Lesley
Sewell, Rod Ismay, David Mason, Malcolm Zack and Gina Gould) and
why was the Zebra Action Summary sent to them? Were they part of the
Post Office litigation team concerned with subpostmaster claims, for
example, or were they management people responsible for deciding
what improvements to make to the Horizon systems, or what?”
659. Amy forwarded Counsel's queries to Rodric Williams (with me in copy),®°? and
we had a call with Rodric Williams and Mark Underwood on 27 July 2018 to
discuss those questions.®° Amy and I relayed POL’s instructions to Counsel in a
call on 31 July 2018.° Counsel did not change their preliminary view that the
Action Summary was not a wholly privileged document. However, they confirmed
that the Project Zebra Desktop Report itself was covered by legal advice and/or
litigation privilege. That was the basis on which we redacted references in the
Action Summary to the Desktop Report (including the word “Zebra” in the
document title) and to Deloitte’s findings and recommendations. Those words
had the capacity to reveal privileged information and so redactions were applied
to them.
660. Following the disclosure of the redacted Action Summary and preparation of the
CIT e-bundle, Freeths wrote to WBD on 22 October 2018 raising concerns as to
862 POL00408730.
®63 POL00255961.
864 WBON0001270.
Page 371 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
whether too wide an approach had been adopted in the redaction of POL’s
documents.®° By way of specific example, they referred to the redactions to
document G/40 in the e-bundle (which was the redacted copy of the Action
Summary) and invited us to reconsider them. They sought confirmation that our
trial Counsel and I (as the person within WBD with overall responsibility for the
disclosure process) had reviewed the redactions and the reasons given for them.
A call was arranged on 23 October 2018 with Owain Draper and Gideon Cohen
(Junior Counsel for the CIT) to discuss this point, who then briefed David
Cavender QC (who was leading them). The Counsel team confirmed that the
redactions were correctly made on the basis that Project Zebra was protected by
legal advice and/or litigation privilege.
661. On 14 November 2018, in the course of the CIT, Patrick Green QC expressed
doubt that the heading of the Action Summary had been properly redacted. He
referred to the fact that the document title within the e-bundle showed that the
word redacted from the title was “Zebra” (this was an inadvertent consequence
of the e-bundle software pulling the title from the document's metadata). We
wrote to Freeths that evening requesting that G/40 was renamed in the trial
bundle to remove the reference to “Zebra’”.®°’ Freeths rejected that proposal.
662. Angela Van den Bogerd was in due course cross-examined on the Action
Summary. In the course of that cross-examination, on 21 November 2018,
Patrick Green QC embarked on a line of questioning that strayed into the
665 WWBON0001339.
66 WWBON0000626.
&87 WBON0001362.
668 WBON0001364.
Page 372 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
outstanding dispute between the parties as to whether “Zebra” in the document
title was correctly redacted:
“Q. I am going to ask you a couple of questions very carefully because
there is a dispute about whether the title of this document is privileged.
So I am not going to ask you what it is. But you know what the title is
don't you, of this document? The redacted word?
A. No. I don’t recall seeing this document, so I can’t—
Q. Are you aware of a project group, named after an animal, which has
been working in relation to issues raised by Second Sight or similar?”
663. David Cavender QC raised an objection to this line of questioning. Mr Justice
Fraser asked Angela Van den Bogerd to step outside the courtroom whilst he
discussed this point with Counsel. Mr Justice Fraser and David Cavender QC
had a brief exchange as to whether the redaction of the single word “Zebra” was
appropriate, during which he requested that David Cavender QC review the basis
for the redactions to the Action Summary:
“MR JUSTICE FRASER. ... I actually made a note to myself, which I was
going to deal with at the end of the witness’s evidence, about this
document because it seems to me the basis of the redactions is going
to have to be reviewed by counsel, not least because it appeared to me,
which now I think is common ground, that the single word is said to be
privileged, and I am struggling with that but—
MR CAVENDER: It depends where you go with it, and as soon as you
unravel a privilege then you get a difficulty.
MR JUSTICE FRASER: I am not sure that is necessarily right but that is
why I am going to invite review by you anyway. It is difficult to grasp how
a single word could be subject to litigation professional privilege or any
of the other types of privilege, so I am going to invite you I think to review
that.”
664. Mr Justice Fraser disallowed this line of questioning on the part of Patrick Green
QC but urged that David Cavender QC review the document and the redaction
of this word afresh. David Cavender QC did so and expressed the view in court
Page 373 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
on 22 November 2018 that he was satisfied that the Project Zebra Report was
privileged and therefore that the redactions to the Zebra Action Report (including
the title of Project Zebra) were properly made.
665. As mentioned above, this was the only document that I specifically recall having
been escalated to me by my team for me to consider redactions. I remain of the
view that the redactions were appropriately applied in order to preserve the
privilege attaching to the Project Zebra Desktop Report and the other documents
produced by Deloitte as part of that investigation. That view was supported by
Counsel and it was further considered and then maintained orally by David
Cavender QC at the Common Issues Trial.
Redactions on grounds of irrelevance and confidentiality
666. Asmall number of the documents disclosed in the group litigation were redacted
not on grounds of privilege, but on the basis of irrelevance and confidentiality.
Redactions on this basis were in the main applied to documents which were
commercially sensitive, typically papers which were submitted to the POL Board,
its General Executive or other committees which operated at that level, and
minutes of the meetings of such bodies. Often these types of documents would
cover a range of different issues affecting POL and which were on the agenda
for a meeting of that body. Of those agenda items, only some were relevant to
the group litigation.
667. Paralegal first level reviewers were not tasked with flagging documents for
potential redaction on grounds of irrelevance and confidentiality (as they were
with documents that were partly privileged or potentially partially privileged). It
was for the qualified lawyers tasked with reviewing documents marked for
Page 374 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
disclosure at the second stage to identify whether redactions ought to be made
on this basis, where the document to be disclosed fell within the abovementioned
category of senior executive documents.
668. The background to this approach being adopted in relation to papers submitted
to the POL Board was that on 14 May 2018, Amy Prime e-mailed me asking “can
we redact the non-disclosable pages [of such documents] — they are littered with
privilege and commercially sensitive”.°©° A later e-mail from Amy records that I
agreed to that approach.§”° Amy later emailed Michael Wharton in my team to
inform him that the process to follow in relation to POL Board papers was to: “...
adopt the same approach as the previous board minutes (ie. redact everything
which is not relevant)’.°"' This was the process which the WBD team then
adopted to POL Board papers. I believe that this was a reasonable approach to
adopt to this category of document because of the heightened confidentiality
and/or commercial sensitivity of the matters contained in them. I also stress that
the approach was to only apply redactions where the information was irrelevant
to the issues in dispute.
669. With reference to Q95.1 of the Request, on 5 March 2019 Freeths wrote to WBD
about redactions applied to documents in the HIT e-bundle and suggested that
31 documents should be reviewed by our trial Counsel.°”? We agreed. As a result
of that review, some redactions were removed and some were maintained. In
relation to those cases where redactions were maintained, Junior Counsel
(Owain Draper) advised that “/ am satisfied that the remaining redactions are
868 WWBONO000155.
870 WBONO000181.
®71 WBONO000176.
872 POL00266947.
Page 375 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
appropriate. I do not say that they could not be challenged, but they are
legitimately made”.*”> Where redactions had been removed, we explained that
this was “for the sake of pragmatic co-operation and to reduce the scope of any
disputes between the parties over the redaction of the documents’; we
maintained our view that the un-redacted material was irrelevant (or at best only
peripherally relevant) to the issues in dispute.°”4
670. During the HIT, Mr Justice Fraser asked Tony Robinson QC to conduct a review
of nine redacted documents. WBD also conducted a review of these documents
and independently decided that, in relation to one of them, the redaction “seems
to have been done on the basis of irrelevance and confidentiality], but this seems
questionable’.®”> This document was de-redacted, and a clean version was
provided to the Claimants. As to the remaining eight documents, Tony Robinson
QC identified that (i) two documents were redacted for privilege and the claim to
privilege was maintained, and (ii) the remaining six documents were redacted
due to irrelevance and confidentiality. In relation to these latter six documents,
Tony Robinson QC deliberately and expressly adopted a broad and generous
approach to what was considered ‘relevant’. On this basis, some of the
redactions were disapplied.
671. In a document review exercise of this scale, different lawyers may fairly form
different views on a document and whether it should or should not be redacted.
It is also inevitably the case that mistakes can happen or misjudgments can be
made, as was the case with the one document where we, on re-review,
873 WBONO0001465.
®74 WBONO0001449.
875 WBON0001574.
Page 376 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
considered a redaction to be ‘questionable’ and Tony Robinson QC agreed.
However, at all material times, my team and I worked very hard to ensure that
disclosure was undertaken properly, efficiently, and proportionately, as is
illustrated by the small number of total redactions which were applied to a large
pool of disclosed documents. Fewer than 500 documents were redacted, of
which 17 had those redactions changed. In a number of these instances this was
driven by my firm's desire to reduce disputes and by Tony Robinson QC’s
deliberately broad approach to the assessment of relevance than anything else.
(v) Documents held by Royal Mail (Q99)
672. By a letter dated 27 February 2019 (POL00003635), WBD informed Freeths that
POL had sought to obtain from Royal Mail pre-2011 audit reports on Horizon
(‘the E&Y Reports”), produced by Ernst & Young Global Limited Liability
Partnership (“E&Y”), but that Royal Mail had expressed concerns about providing
those reports without a court order. By a later letter dated 15 March 2019
(POL00003570), we informed Freeths that this was incorrect and apologised for
the error. I am asked by Q99 what the basis for making the original (erroneous)
statement was, and how I discovered the mistake.
673. The background to these matters is as follows.
674. Audit reports on Horizon produced by E&Y from 2011 onwards were disclosed
by Post Office to the Claimants in August 2018. On 19 September 2018, Freeths
wrote to WBD requesting disclosure of all audit reports on Horizon that had been
produced since the introduction of Horizon (whether produced by E&Y or
Page 377 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
others).°”° On 3 October 2018, we responded to Freeths stating that we were
taking steps to locate and disclose the reports referred to.5””
675. There was then a separate line of enquiry, beginning in October 2018. Lucy
Bremner of WBD emailed Mark Hotson of POL on 24 October 2018. She
informed him that WBD was in the process of drafting witness statements for the
HIT, and explained that the Claimants’ IT expert, Jason Coyne, had referred to
an audit report produced by E&Y in 2011 which identified problems with the
Horizon Credence application. She explained that WBD needed to understand if
the comments made by Mr Coyne were correct and whether any changes were
made thereafter.5’°
676. On 26 October 2018, POL’s Head of Internal Audit, Johann Appel, wrote to
Angela Van-Den-Bogerd (copying Lucy), informing her that he was planning on
contacting the company secretarial administrative department of POL (“Co-Sec”)
to see if the papers and minutes of the Audit Risk and Compliance Committee
(‘ARC Committee”) provided more information.6”? I was not copied into this
email.
677. On 31 October 2018, Rebecca Reay (a company secretarial administrator at
POL) wrote to Royal Mail to seek access to the minutes from the Royal Mail
Group Audit, Risk and Compliance meetings from 2010 and 2011.5°° No-one
from WBD was initially copied into this email.
876 POL00257537.
877 POL00285759.
678 WBONO0000198.
®78 WBONO0000283
680 WBON0001492.
Page 378 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
678. Johann Appel provided Lucy with an update on 2 November 2018. I was copied
into this email thread, which included Johann Appel’s email of 26 October 2018
(where he wrote that he was searching for the ARC Committee papers and
minutes). In his update of 2 November 2018, Johann Appel wrote that POL: “were
unable to locate the relevant documents prior to 2011 at Post Office” and had
therefore asked Royal Mail to look for the documents in their archives. He said
that POL were chasing daily for a response.
679. On 6 November 2018, Johann Appel reported (to Angela Van-Den-Bogerd and
Lucy Bremner) that POL was still waiting for Royal Mail to search their archives
for “documents related to the FY2010 ARC and EY report’. He noted that the
2011 E&Y Report referred to the Credence issue only as part of an update on
findings from an earlier 2010 audit. He said that he would therefore continue to
chase Royal Mail to find the “original 2010 EY report’ and the “2010 ARC
minutes” .°°2
680. On 9 November 2018, Johann Appel reported back to Lucy to update her on
POL’s lack of progress on this issue. He explained that whilst Royal Mail's
archive personnel had initially appeared willing to assist, the matter was then
referred Royal Mail’s legal department, who were unwilling to release any
documents to POL.®®°
681. Later on 9 November 2018, Rodric Williams emailed Lucy Bremner and Johann
Appel, copying me. Rodric Williams wrote that he had spoken to the legal
681 WBONO0000282.
682 WWBONO001487
683 POL00042127.
Page 379 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
department of Royal Mail and asked Lucy to call him.®* I cannot recall whether
that call was made by Lucy.
682. On 12 November 2018, Rodric Williams then emailed the Royal Mail legal
department in accordance with a request made by Lucy Bremner. He explained
that POL were seeking to locate a “2010 Ernst & Young Management Letter’ and
“a Royal Mail Group Audit Committee or Audit, Risk & Compliance Committee
(ARC) minutes for 2010/2011”.°85 I was not copied into this email itself, though I
was copied into an email setting out a draft of it®°° and the email as sent was
forwarded to me on the same day.
683. It appeared from an internal POL email sent by Johann Appel to Rodric Williams
on 14 November 2018 that Royal Mail proposed that, whilst extracts from the
documents in question could be read to POL, Royal Mail was unwilling to release
copies of the documents without a request through legal channels.°°”
684. On 8 February 2019, Freeths asked WBD to confirm that all audit reports on
Horizon had been disclosed.®° In response, on 11 February 2019, we informed
Freeths that a search for audit reports had been undertaken and the results
would be disclosed. Further, we explained that enquiries were being made to
understand who was appointed as auditor of Horizon prior to 2011.°89 On 12
February 2019, we wrote to Freeths and listed a number of audit reports which
were being disclosed.
684 WBON0001359.
685 WBON0001360.
686 WWBONO0001693.
687 WBON0001485.
688 WWBONO001412.
688 POL00263874.
690 WBON0001416.
Page 380 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
685. On 12 February 2019, Johann Appel confirmed to Lucy Bremner (in an email
copied to me and Amy Prime) that E&Y were the Horizon auditors prior to 2011."
686. On 13 February 2019, Lucy responded to Johann Appel’s email and asked him
to confirm that “all audit reports pre-2011 are in Royal Mail’s possession and not
yours” (emphasis added).°9
687. Johann Appel responded the same day (again, copied to me and Amy Prime)
stating: “The solution that Rod[ric Williams] agreed with Royal Mail’s lawyer was
that they would verbally share any relevant extract from those audit reports and
if Rod thinks it is applicable, then he would have to subpoena them to release
the reports to Post Office” (emphasis added).®® I did not consider this approach
to be surprising because the voluntary release of documents can raise
confidentiality and data protection issues. As such, it was understandably safer
for Royal Mail to be ordered to disclose the documents, because then they would
be protected in the event that such documents did raise confidentiality and data
protection issues.
688. On the basis of this email, I understood that the arrangement which had been
made as between Royal Mail and POL (which required a subpoena to be made)
related to the pre-2011 audit reports that we had been requesting. In turn, I
understood that the pre-2011 audit reports had been requested from Royal Mail.
689. That this was my (mis)understanding at the time was reflected in the steps which
we took after receiving this email:
691 WBONO0000170.
682 WBON0000173.
&°3¥VBONO0000167.
Page 381 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
689.1.On the basis of these instructions, WBD then wrote the letter dated 27
February 2019 (i.e. POL00003635), and informed Freeths that POL had
requested the pre-2011 audit reports from Royal Mail but that Royal Mail
had expressed concerns about providing the reports without a formal order
for third-party disclosure.
689.2.0n 13 March 2019, following receipt of a letter that day from Freeths®*
Amy Prime emailed me to say that her plan was for POL to call Royal
Mail’s legal team to confirm that Royal Mail were refusing to disclose the
E&Y Reports and to warn them that WBD would seek a third-party
disclosure order. Amy suggested that once this had been done, a letter
could be sent to Freeths to confirm.5® The draft letter stated that the pre-
2011 reports were not within POL’s control, and it may be necessary to
seek an order from the court for disclosure of the documents.®°
689.3.Then, on 14 March 2019 (which was Day 4 of the HIT), Patrick Green QC
informed Mr Justice Fraser that WBD had said that Royal Mail were
concerned about providing the pre-2011 E&Y Reports without a court
order. Tony Robinson QC informed Mr Justice Fraser that the previous
year, POL had asked Royal Mail for the “documents” and that Royal Mail
had said that they would not provide them voluntarily and a court order
would be required. The statement made by Tony Robinson QC reflected
my understanding of the matter at that time (as WBD had set out in its
letter to Freeths dated 27 February 2019).
64 WBONO0001472.
£85 WWBONO000320.
686 WBON0000322.
Page 382 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
689.4.Mr Justice Fraser therefore directed the Claimants to issue an application
against Royal Mail for third-party disclosure in respect of the E&Y Reports.
689.5.That same day, Amy Prime sent POL a draft email to send on to Royal
Mail, asking Royal Mail to reconsider its position that it would not be
prepared to provide copies of the pre-2011 E&Y Reports without a Court
order.®°” This draft email reflected what we continued to understand was
the position, namely, that in 2018 POL had sought copies of the pre-2011
E&Y Reports but that Royal Mail refused to provide those documents
without a formal court order. The draft email asked Royal Mail to
reconsider its position.
689.6. That day, 14 March 2019, Rodric Williams sent an email to Luke Ryan of
Royal Mail along the lines of the draft provided to him by Amy. WBD
also wrote to Royal Mail to ask them to provide copies of the pre-2011
audit reports.®°9
689.7.On 15 March 2019, in a telephone call, Luke Ryan of Royal Mail informed
me that POL had never previously requested the pre-2011 audit reports.
This was the first time that I became aware of the misunderstanding that
had arisen.
690. A series of emails (evidencing the original request for documents made by POL
to Royal Mail) were then forwarded to me on 15 March 2019.7°° What had
actually happened, I could now see, was that POL had previously only requested
687 WBON0001483.
688 WBON0000201.
688 POL00022691
700 WBONO0001488; WBONO0001492; POL00042127; WBON0001491; WBON0001487.
Page 383 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
the “2010 E&Y Management Letter ... and/or the ARC minutes”.’°' However, in
light of Johann Appel’s email of 13 February 2019 and the thread which preceded
it, we had mistakenly understood that POL had actually requested the pre-2011
audit reports.
691. I immediately telephoned James Hartley of Freeths to inform him following my
discovery of the mistake. I promised to confirm what had happened in writing,
and at around 12:30 on 15 March 2019, we wrote to Freeths to correct the
position with respect to the pre-2011 E&Y Reports. I made plain that the error
was ours and that we sincerely apologised for it.”
692. On 18 March 2019, Royal Mail provided Freeths with the audit reports from 2008,
2010 and 2011.7°3
693. That same day, Mr Justice Fraser ordered that POL to file a witness statement
by 21 March 2019, setting out steps it had taken to obtain the pre-2011 E&Y
Reports, details of the requests made prior to 18 March 2019 to obtain the reports
from Royal Mail and E&Y and, if no requests were made, to explain why this was
the case. I filed my thirteenth witness statement on 21 March 2019 setting out
the history of what had happened (insofar as that could be done without a waiver
of privilege).
694. As the contemporaneous documents demonstrate, this was a misunderstanding
whereby I and others at WBD thought, in light of our instructions, that Royal Mail
had been asked for the pre-2011 EY Reports and had responded that a
“subpoena” would be required to provide them, when in fact Royal Mail had been
702 POL00269022
703 WWBONO001510
704 POL00269053
Page 384 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
asked for closely connected but different audit-related documents. As soon as I
became aware of the misunderstanding, I contacted James Hartley at Freeths to
explain the position. I immediately confirmed the position in writing and
apologised for it. I was also at pains to ensure that our Leading Counsel
apologised to the Court.
O. PREPARATION FOR THE COMMON ISSUES TRIAL (Q90, Q92 to Q94)
695. In this section I respond to the Inquiry’s questions about my involvement in the
preparatory work for the Common Issues Trial or CIT. These are set out at Q90
to Q94 of the Request, although Q90.1 and Q91 are addressed above in Section
N on disclosure, at §§521 ff.
696. Before turning to those questions, there is an overarching topic that I consider
relevant both to POL’s preparation for the CIT, and also to its response to the
resulting Common Issues Judgment including the appeal against that judgment
and the Recusal Application (which I deal with below in Section Q). That topic
relates to the scope of the evidence submitted by the Claimants for the CIT and
the consequences of this for how POL prepared its case.
(i) The scope of the evidence for the CIT
697. In broad terms, the CIT was intended to resolve 23 preliminary issues relating to
POL's relationship with its SPMs, the effect of which was heavily contested
between the parties. This included construing the express terms of the SPMC
and NTC, deciding whether any terms should be implied (and, if so, what terms
and on what basis), and determining issues of incorporation of allegedly onerous
Page 385 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
terms as well as the application of the Unfair Contract Terms Act 1977. The CIT
expressly did not extend to consideration of whether POL had breached of the
terms of its agreements with the Claimants or indeed any other duties (for
instance, in tort) which it may owe SPMs. Thus, for example, whether POL was
subject to an implied term requiring it to ensure that the Horizon IT system was
reasonably fit for purpose with within the scope of the CIT; but whether there
were in fact any problems with Horizon was not. Similarly, whether POL owed a
duty to provide adequate training was with the scope of the CIT, but the quality
of the training actually provided to the Claimants was not.
698. Without wishing to oversimplify these matters, my view at the time of the group
litigation was that such questions of construction, incorporation and implication
fell to be determined against the background of the objective facts that were
known to both parties at the time of entering the contract. This ‘factual matrix’, as
I understood it, generally excluded (for example) the parties’ subjective intentions
in agreeing to the contract; what each of them said in the course of pre-
contractual negotiations; and matters that became known to them only after
entering into the contract, such as how each of them performed in practice. Since
matters which did not form part of the legitimate factual matrix were irrelevant for
the purposes of identifying and interpreting the contractual relationship between
two parties, my expectation was that evidence of such matters would not be
admissible at the CIT (which was a trial to determine the contractual framework
in place between POL and SPMs).
699. As I have indicated, these matters were not straightforward. First, there were a
limited number of issues where there could be narrow exceptions to the position
outlined above; and second, the precise boundaries of the legitimate factual
Page 386 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
matrix were not always clear-cut. For example, where the Claimants alleged that
the suspension and termination provisions were a sham, evidence of those
provisions being used in practice (consistently or not with the express written
terms) may be relevant to ascertaining the ‘true’ agreement in in relation to those
terms. Further, the Claimants were advancing a case that the extent and
application of common law agency principles turned on the facts of the SPM-POL
relationship and were also asserting that, contrary to the express contractual
terms, POL was acting as an agent for SPMs.
700. My understanding was shared by the Counsel team on this matter. We became
concerned at an early stage in the litigation that the factual matrix on which the
Claimants were seeking to rely was far wider than would be admissible on any
conventional understanding of the law, even allowing for exceptions of the kind
referred to above. The GPOC served on 23 March 2017 contained a section
headed ‘factual matrix’, with a few brief paragraphs (41 to 45) which referred in
general terms to there being an imbalance of power between POL and SPMs,
and the relationship between them being akin to an employment relationship,
and with a further paragraph (46) which simply read:
"The Claimants will rely on these and other aspects of the factual matrix
as particularised elsewhere in these [GPOC] and as may further be
established as relevant in individual cases".
701. That did not assist us in understanding the facts the Claimants considered were
relevant to determining the relationship between the parties. Accordingly, on 27
April 2017, POL made a CPR 18 request in respect of paragraph 46 of the GPOC,
asking the Claimants to clarify the aspects of the factual matrix they were relying
on for the purposes of construing the pleaded agreements. Their response dated
16 May 2017 was as follows:
Page 387 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
“8. Not entitled, since the Claimants’ case overall is sufficiently pleaded.
Without prejudice to that:
8.1. As to (a), all facts pleaded, including those at paragraphs 9, 12-39,
41-45 and 81.
8.2. As to (b), as pleaded, namely, all facts as may further be established
as relevant in individual cases (following SOls, disclosure and individual
Particulars of Claim in test cases). The Defendant is reminded that these
proceedings are managed under a GLO.”7
702. My view, and that of the Counsel team, was that this did not help us to understand
the Claimants' intended approach, but that it was apparent that they wanted to
rely on matters which would never normally be considered admissible factual
matrix. For example, paragraphs 12-39 of the GPOC contained substantially the
whole of the Claimants’ case on the reliability of the Horizon system, remote
access, the adequacy of the training with which they had been provided, and the
sufficiency of POL’s investigations into shortfalls in their accounts. In other words,
these were clearly matters that went to the adequacy of POL’s performance of
its contractual duties, breach, and causation of loss.
703. Therefore, at the first CMC in October 2017 POL sought and obtained an order
that there should be an agreed statement of factual matrix before witness
statements for the CIT were exchanged (the “factual matrix document”). The
intention was that this would give the parties the ability to resolve any issues on
what evidence would be admissible prior to witness evidence being served. The
date set for the factual matrix document was 29 June 2018; as I mentioned above
(§318), witness statements for the CIT were to be filed and served on 11 August
2018.
705 WBONO001682.
Page 388 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
704. Following the CMC, we set out our concerns about the Claimants’ response to
our CPR 18 request in a letter dated 1 November 2017.” Our letter set out our
view that the approach the Claimants seemingly intended to adopt was wrong in
law, and also highlighted that any attempt to rely on large amounts of
inadmissible evidence would likely cause significant cost and disruption to the
CIT timetable. As part of this exchange, we encouraged Freeths to engage at an
early stage in order to agree the factual matrix document.
705. Freeths responded on 9 November 2017 and it was clear from their letter that
the Claimants did indeed intend to rely, as part of the factual matrix for the
Common Issues, on post-contractual matters relating to performance and breach
(including the “existence of software coding bugs and errors” in Horizon, and the
“factual circumstances in which [Lead Claimants] signed Branch Trading
Statements’). Although Freeths acknowledged that the CIT was not to determine
whether POL had breached its contractual duties, they said that “the Court may
well have to consider what the consequences of a breach might have been and
in what circumstances [SPMs] might have found themselves in breach on the
parties’ various constructions of the contracts”.’°”
706. We did not agree with this analysis; if correct, it would mean that evidence going
to performance, breach, and the impact thereof on individual contract parties
would always be admissible factual matrix, which was not the state of the law as
we understood it. There were then CMCs on 2 February 2018 and 22 February
2018 where this issue was discussed at length with Mr Justice Fraser in the
context of the scope of disclosure orders for the CIT. My overall impression from
706 WWBONO001220
707 WBON0001219
Page 389 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
those hearings was that the Mr Justice Fraser broadly agreed with POL's
narrower view of the admissibility of evidence at the CIT.”°8
707. We continued to correspond with Freeths on this issue and on 2 March 2018 the
Claimants provided their draft proposed factual matrix document.’ This
confirmed rather than allayed our concerns. Examples of the factual matters on
which the Claimants proposed to rely as relevant to the Common Issues included
the adequacy of the advice provided by the NBSC Helpline, evidence of errors
in Horizon, and the fact that POL had pursued individual Claimants for shortfalls.
We wrote to the Claimants setting out our concerns on 20 March 2018.7'°
708. The Claimants’ apparent determination to rely on matters going beyond what we
believed was relevant to the Common Issues put POL in a difficult position. It
seemed unlikely that the parties would be able to arrive at an agreed factual
matrix document by the end of June 2018 as directed. Seeking directions
designed to keep all of the Claimant's proposed inadmissible material out before
the service of witness statements could give Mr Justice Fraser the wrong
impression that the merits of the Common Issues lay with the Claimants. In any
event, this approach was unlikely to fully succeed. We were therefore likely to be
faced with a situation where the Claimants would be serving at least some
evidence the Counsel team and I considered inadmissible and irrelevant (but
potentially prejudicial), and POL would either have to address those contentious
matters in its own evidence notwithstanding our principled position that they were
708 WWBON0001337. See for example page 9, Line H, where Mr Justice Fraser said to the
Claimants’ Counsel, "/ have your point about the orthodox factual matrix. Mr. Cavender, to be
fair to him, explained this very clearly on the last occasion. We did not go into the authorities
but I thought I made it clear that his approach on construction was indeed correct"
709 POLO01 10872
710 POL00363651
Page 390 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
irrelevant, or POL could ignore them in its own witness evidence but in doing so
would risk leaving those points unanswered.
709. On 28 March 2018, we produced a briefing paper for the Steering Group
forewarning of this issue.”'? The paper highlighted that we could not predict the
extent to which Mr Justice Fraser would allow the Claimants to serve evidence
we believed to be inadmissible, and recommended that POL should “not ...
address the inadmissible evidence in the witness statements we are currently
preparing’.
710. David Cavender QC further raised our concerns at the CMC which took place on
5 June 2018. For example, the following discussions took place (I have
underlined what I believe are the key points):712
“MR. JUSTICE FRASER: No, let me deal with it on that basis. Whatever
the factual evidence upon which you seek to rely it has to be relevant to
the Common Issues.
MR. GREEN: Correct.
MR. JUSTICE FRASER: If it is not relevant to the Common Issues it is
not admissible.
MR. GREEN: Absolutely right, there is no dispute, to [sic] uncertainty
about that.
MR. JUSTICE FRASER: In those circumstances it is difficult based on
reading the authorities to see for example, to use Mr. Cavender's
example, how evidence of breach could remotely be relevant to the
Common Issues Trial.
MR. GREEN: We have at some length sought to explain that in
correspondence.
MR. JUSTICE FRASER: Would you like to explain it to me?
MR. GREEN: Certainly.
711 POL00006408.
72 POLO0041899.
Page 391 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
MR. JUSTICE FRASER: Perhaps not at some length but just relatively
succinctly.
MR. GREEN: The characterization of matters being matters that go to
breach is the defendant's characterisation of those matters.
MR. JUSTICE FRASER: I do not understand that submission for a
moment I am afraid simply as a matter of English.
MR. GREEN: Someone says, "I was provided with this training which I
found inadequate and it did not help me do X", let us assume that is
going to be the evidence. Now, my learned friend says that is evidence
that goes to breach, but that is wrong analytically.
MR. JUSTICE FRASER: Show me which Common Issues it would go to.
MR. GREEN: May I take it in stages?
MR. JUSTICE FRASER: Yes.
MR. GREEN: The first point is that on Common Issues number 1,
relational contract the court has to decide that by looking at the nature of
the contract.
MR. JUSTICE FRASER: Correct.
MR. GREEN: As in fact it worked in practice to see whether or not it was
a contract which requires the parties ----
MR. JUSTICE FRASER: I do not think one looks at the nature of the
contract as it worked in practice. One looks at the nature of the
relationship between the parties to the contract to see if the necessary
ingredients, or if there are any new ones which have not yet been subject
to authority. Whatever the necessary ingredients are for a relational
contract are, print [sic] or not.
MR. GREEN: Correct. Then we reformulate it to say, was the contract
one which in practice required the fair dealing and good faith requirement
et cetera in the ----
MR. JUSTICE FRASER: That does not require breach.
MR. GREEN: No, but, my Lord, my learned friend has captured the
language, we say, quite wrongly. There are two points, contractual
orthodoxy from which we do not depart at all. The first point is that when
you are looking at the construction of a contract you look only at the
evidence as it was when the parties contracted. We are not going to
invite your Lordship to look at any evidence after the parties contracted
to construe the agreement that they entered into on that date.
MR. JUSTICE FRASER: Good, because that would be inadmissible.
Page 392 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
MR. GREEN: Of course. I am trying to clear the ground where the dispute
is.
MR. JUSTICE FRASER: By definition the breach must happen after the
contract ----
MR. GREEN: Of course, we are not talking about ----
MR. JUSTICE FRASER: My question to you was predicated specifically
by reference to breach.
MR. GREEN: Breach assumes one has identified what the legal
obligation is first which we have not even done, that is what the Common
Issues Trial is about. My learned friend's characterisation is speculative.
MR. JUSTICE FRASER: I will tell what you I am going to do about this
because I am have grave difficulty in following it, but it is also
undoubtedly the case that there are bear traps left, right and centre in
my attempting to identify in advance ----
MR. GREEN: Precisely.
MR. JUSTICE FRASER: ---- when you can and cannot do in your
evidence. So this is what I am going to do. I am going to express myself
very clearly. If you serve evidence of fact which includes passages which
are plainly not relevant and, hence, not admissible, Mr. Cavender is
going to have a choice. He can either simply say, "I am not going to be
cross-examining at all" or he is going to issue an application to have it
struck out. If he does issue an application to have it struck out and that
application is effective, it will involve the court going through it and simply
striking out large amounts. The court will make time to do that but
cringing costs consequences will follow.
MR. JUSTICE FRASER: It is an exercise which will be very tedious and
expensive and it will take a day or two but it can be done.
MR. GREEN: Yes. My Lord, we expect all of that. That is what we expect
but we also note that my learned friend having initially opposed this point
conceded it before you in the transcript, we can find a reference if you
want, that each time Post Office exercises its entitlement to vary the
contractual relationship or the contractual obligations of
Subpostmasters, that falls to be construed as the position is known to
the parties at that time.
MR. JUSTICE FRASER: Of course, that is contractual orthodoxy.
MR. GREEN: Precisely, that is all ----
MR. JUSTICE FRASER: But it does not open the door and it might be
that this is all a concern without any real substance. It does not open the
Page 393 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
door to wide-ranging evidence of fact which appears to be Post Office's
concern, that cannot possibly form part of the factual matrix.
MR. GREEN: Precisely. We have taken that on board, I hope.
MR. JUSTICE FRASER: I know, you always do say you take it on board
and you all say that you are following contractual orthodoxy. It might be
that you are.
MR. GREEN: I am grateful.
MR. JUSTICE FRASER: At the moment, without the documents in front
of you to be able to look at it with any sort of concrete analysis, it is
difficult for me to do any more. To continue the quasi military analogies
from earlier this afternoon, a very powerful shot has now been fired
across your bows on two occasions and I do not mean by Mr. Cavender
or Mr. De Garr Robinson; I mean by me.”
711. My view following that hearing was that Mr Justice Fraser continued to broadly
agree with POL's position on what evidence would and would not be admissible
in relation to the Common Issues. However, given that he had indicated that
these matters fell to be dealt with after service of the evidence, I remained
concerned that the Claimants were going to submit evidence that went beyond
the matters known to the parties at the point of contracting.
712. On 17 July 2018, I presented a Decision Paper to the Steering Group on whether
or not POL’s witness evidence should deal with matters which we anticipated the
Claimants would rely on, but which we considered to be irrelevant and
inadmissible.’' This paper drew on discussions I had had with Counsel on this
point and reflected their views as well as my own. Looking at that paper again
now I note that:
712.1.We said that “(with a high degree of confidence)... the orthodox legal
position is that the only admissible evidence is that which was known to
713 POL00358103
Page 394 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
the parties at the time of entering into the contract.” That reflects my
understanding of the legal position, as I have outlined above.
712.2.We believed “that the Judge [was] with us given comments he made at
previous hearings’. That is demonstrated by the parts of the transcripts I
have excerpted above.
712.3.We also outlined a number of exceptions of the kind I have alluded to
above, including (but not limited to): (i) where a contract was varied,
knowledge at the time of variation could be taken into account for the
purpose of construing the varied terms; (ii) we wanted Mr Justice Fraser
to have a basic understanding of how POL operated, which would involve
describing some aspects of how branches were run; (iii) countering some
of the Claimants’ implied terms arguments might require evidence on the
practical consequences of those terms; and (iv) POL might need to present
some real-world evidence on the fairness of the terms that the Claimants
alleged were unfair.
713. The Decision Paper outlined three options on how POL should deal with the
Claimants’ allegations ranging from responding to all of them to not responding
to any allegations that we said were inadmissible whatsoever. The paper
concluded that both extremes carried too much risk and that POL should go with
a middle ground of responding selectively but with a strong bias towards not
responding to anything thought to be inadmissible. In particular, rather than
responding to allegations relating specifically to Lead Claimants, Counsel and I
recommended that POL should put forward some high-level generic evidence on
how its business worked and performed in practice. In practice, I recall that the
final statements produced by POL took an even tighter line on admissibility and
Page 395 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
were very largely constrained to pre-contractual matters. See below, §§979,
where I provide further observations on this.
714. On 19 July 2018, the parties’ exchanged further letters setting out their respective
positions.’"4 In our letter, we “[put] down again the following marker: anything that
was not known (or at least knowable) to a person in the position(s) of the parties
at the time of contracting cannot be admissible matrix of fact ... It follows that
nothing which happened after entry into the relevant contracts can be admissible
evidence for the purposes of the Common Issues trial’. We identified three
narrow possible exceptions to the rule, as follows:
715. “a. Your clients' case is that the parties' ‘true agreement as to
termination was made manifest after the relevant contracts were entered
into. In principle, evidence going to this could be admissible — although
it remains unclear ... what that evidence could be.
716. b. Insofar as any contract term was varied, the factual matrix
relevant to construing that term, as varied, will be that obtaining at the
time of variation. That is the proposition set out, for example, at
paragraph 4.4 of Mr Bates' Reply. Post Office agrees. However, for this
exception to be relevant the Claimants (or any of them) would have had
to (a) plead a relevant variation, (b) set out the respects in which the
factual matrix existing at the time that the variation was made was
different from that existing at the time of entry into the contract, and (c)
set out the way in which those differences in factual matrix are said to
have a bearing on construction of the varied term. They have not done
so.
717. ¢. Some, very limited, evidence of post-contractual behaviour may
be relevant to the question of agency. It is, however, difficult to see how
individual evidence from particular claimants (as opposed to general
evidence of practices persisting over time) could assist.”
718. Since it was apparent that the Claimants’ continued intention was to serve
evidence going beyond the scope even of the exceptions outlined above, we
714 POL00255849; POL00255848
Page 396 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
notified them that we intended to invite Mr Justice Fraser to set aside time in
September, to enable him to (if necessary) go through the Claimants’ evidence
and strike out those parts which were inadmissible and/or irrelevant. We
reminded Freeths that “this was the course of action mooted by the Learned
Judge at the hearing on 5 June 2018”. I believe we then wrote to Mr Justice
Fraser advising him of the likelihood that an application would be forthcoming.’‘®
719. When the parties exchanged their witness evidence on 24 August 2018, the
Claimants’ evidence contained lengthy sections dealing with what we considered
to be inadmissible matters, including: complaints in relation to the Lead
Claimants’ training; the operation of the Helpline; problems the Lead Claimants
said they encountered with Horizon; POL’s audits of their branches;
investigations into shortfalls; and events surrounding the Lead Claimants’
suspension and termination. My opinion was that these were properly matters
for a future breach trial and not the CIT.
720. Having reviewed the Claimants’ evidence, we wrote to Freeths on 31 August
20187'® to see if this issue could be resolved between the parties but,
unfortunately, it was not possible to do so. As the next case management hearing
was listed for 19 September 2018 (with the CIT due to start on 7 November
2018), it was necessary to act promptly and so on 5 September 2018, POL issued
the application to which Mr Justice Fraser had alluded at the 5 June 2018 CMC,
namely, to strike out inadmissible parts of the Claimants’ witness evidence (the
“strike out application”). The application did not seek to strike out all of the
715 I have not been able to locate a copy of the relevant correspondence, but it is referred to in the
transcript of the hearing on 11 September 2018 (referred to at §722 below, at p.356F-G. It is
also referred to at paragraph 1.1 of the 13 September 2018 Steering Group Decision Paper,
referred to below at §722: POL00006455.
746 WBON0001313
Page 397 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
evidence which we considered to be inadmissible, but those parts we identified
as clearly and obviously crossing the line.
721. Ahead of this application, we had submitted a Decision Paper to the Steering
Group reflecting WBD’s and Counsel's joint view that the application should be
made, together with, inter alia: (i) a note of advice the Counsel team had
previously prepared on what evidence would be admissible at the CIT, (ii) a copy
of my draft ninth witness statement in support of the proposed strike out
application, and (iii) marked up copies of the Lead Claimants’ witness statements
highlighting the passages targeted by the proposed application.’"” The Decision
Paper explained that:
721.1.It was unusual to seek to strike out evidence in advance of a trial, and
judges were generally reluctant to do so. Even so, the proposed
application was justified in the circumstances because of (i) the proportion
of the Claimants’ evidence which we thought was irrelevant and
inadmissible, and (ii) the extent to which that material overstepped the line.
We recommended a balanced approach whereby POL would “limit the
strike out application to that material which we confidently believe to be
inadmissible”. Although there was “more material in the evidence that is
arguably also inadmissible ... we do not propose that Post Office attempts
to strike this out before trial’.
721.2.We did not consider it advisable to attempt to deal with the admissibility
issues at trial. There may not be room in the trial timetable for this, and in
any event waiting until trial to raise the admissibility points would
717 Decision Paper: POL00256731; Counsel team’s advice: POL00256627; draft ninth witness
statement: POL00256583.
Page 398 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
significantly complicate Counsel's trial preparation and would give POL no
opportunity to take mitigating action (for example, by preparing further
evidence) if Mr Justice Fraser positively decided that some or all of the
Claimants’ evidence was relevant to the Common Issues.
721.3.The Claimants’ motivation appeared to be to “prejudice the Judge's
thinking into seeing them as a vulnerable group of quasi-employees who
need to be protected from Post Office and thus hope that he will interpret
the contract terms in a way that is favourable to the Claimants”. Seeking
to strike out their evidence that went to matters of performance and breach
in advance of trial would mitigate this risk. The risk of prejudice was a
legitimate consideration to draw to POL’s attention, however the central
reason for the proposed application was that the evidence in question was,
in our view, irrelevant and inadmissible. As the Decision Paper made clear,
this evidence would be admissible at any future trial on liability, but we did
not believe that it was ripe for consideration at this juncture.
721.4.There was a risk that POL would be seen to be acting “oppressively’ in
issuing the application, however WBD and Counsel considered this to be
“unlikely given (i) [Mr Justice Fraser’s] previous comments and (ii) this is
an important substantive issue”.
722. The Steering Group approved the proposed application and we initially sought to
have it dealt with at the hearing on 19 September 2018. In the event, due to an
administrative error by the Court, the strike out application did not come to Mr
Justice Fraser's attention until around 10 September 2018; he listed a mention
hearing the following day at which he gave directions for the application to be
dealt with at a hearing on 10 October 2018. Since this meant the application
Page 399 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
being dealt with much closer to trial than we had originally envisaged, we
reconsidered the approach. On 13 September 2018, a further Decision Paper
was submitted to the Steering Group which set out that “/tJhe view of the legal
team (including both QCs) is that the application should not be withdrawn”, but
that we thought POL should seek to adjourn the application to be dealt with at
the CIT.”'® This was because:
722.1.At the mention hearing on 11 September 2018, Mr Justice Fraser
appeared to still be “broadly with Post Office” on the substance of the
admissibility issues. However, he had expressed reservations about
striking out evidence in advance of the trial as being the correct procedural
approach. We advised that: "In light of the Judge's comments, we believe
that application will, more likely than not, fail [...]— not on merit, but due to
the Judge's approach to the procedural management of this litigation."
722.2.We advised that the application would have the best prospects of
succeeding, if the application was dealt with at trial, as by then the Judge
would have read all of the evidence and would be fully prepared to deal
with the question of admissibility.
722.3.There was little point (from POL’s perspective) in having the application
heard on 10 October 2018 given the proximity of that date to the start of
the trial (especially as Mr Justice Fraser’s judgment on the application
might not be available until after the CIT concluded).
718 POLO0006455.
Page 400 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
722.4.The trial timetable had now been set and Fridays were reserved for
applications, so there was now some leeway within the timetable to deal
with the strike out application at the CIT itself.
723. In line with that advice, at the hearing which took place on 19 September 2018
Mr Justice Fraser was invited to postpone the strike out application. He declined
to do so, and the application therefore went ahead on 10 October. It was
ultimately by Mr Justice Fraser in his judgment [2018] EWHC 2698 (QB), handed
down on 17 October 2018 (the “Strike Out Judgment”). In short, Mr Justice
Fraser considered that the evidence targeted by the application was potentially
relevant and POL had therefore “not satisfied the necessary test to have these
passages struck out’ at the pre-trial stage: [53]. At the same time, he
acknowledged that this decision did not mean he could go on to rely on matters
which were outside the proper factual matrix when determining the Common
Issues; as he observed at [53]: “should I in the fullness of time make findings on
the Common Issues by taking into account matters irrelevant in law (and hence
inadmissible) on some of those Common Issues, there is a remedy available”.
This was therefore the footing on which the parties proceeded to trial.
724. I appreciate that this section goes into detail on points that the Inquiry has not
directly asked questions about. However, I believe that understanding the basis
of POL’s objection to the evidence served by the Claimants for the CIT, together
with the chronology outlined above, is key to understanding some of the issues
the Inquiry has identified. In particular, it had a critical bearing on how POL’s
evidence was prepared for the CIT, which I expand upon further in response to
the Inquiry'’s Q90.2 and Q92.4 below. Further, the fact that the Claimants’
evidence was allowed into the CIT in its totality influenced the way in which cross-
Page 401 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
examination was approached, as I set out below in response to Q90.3. It also
had a significant bearing on the events which followed the handing down of the
Common Issues Judgment, in that that judgment dealt extensively with the
factual matters relied upon by the Claimants, in a manner which I — together with
the rest of POL’s legal team — considered to be unfair in the circumstances and
incorrect as a matter of law (and indeed, contrary to the indications that had been
given in the Strike Out Judgment). These factors ultimately led to POL’s appeal
against the Common Issues Judgment and the Recusal Application, as I explain
further in Section Q below.
725. Overall, whilst recognising that there were some narrow exceptions, I was of the
view that it was reasonable for POL to adopt the stance that large parts of
evidence tendered by the Claimants were not relevant to the Common Issues.
Having made that general observation, I turn to the specific aspects of
preparation for the CIT that the Inquiry has asked me about at Q90 and Q92 to
Q94 of the Request. I deal first with Mr Justice Fraser’s Strike Out Judgment,
before turning to POL’s preparation of its substantive case for the CIT.
(ii) The Strike Out Judgment (Q94)
726. By Q94 I am asked to comment on the Strike Out Judgment, which dismissed
POL’s pre-trial application to strike out parts of the Claimants’ evidence. I have
described the background to that application in the preceding section, where I
have also dealt with the legal team’s advice on the merits of the application. Here,
I deal with the Inquiry’s questions about the criticisms of POL’s conduct contained
in the Strike Out Judgment.
Page 402 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
727. I identify the main points of criticism made by Mr Justice Fraser in relation to
POL’s conduct as follows:
727.1.At [6], Mr Justice Fraser said that “the making of a GLO at all was opposed
by the defendant’: [6]. He was mistaken on this point. As I have explained
above (for example at §313, §350, §387), POL had in fact agreed in
principle to the making of a GLO from the outset. Advice was received from
Tony Robinson QC to this effect as early as 9 June 2016 (see §§426-427),
and POL communicated its agreement in principle to a GLO to the
Claimants in (inter alia) its LOR. There was some disagreement as to the
terms of the GLO, and whether and to what extent the cut-off date should
be extended. These points were resolved at the GLO Hearing and first
CMC in October 2017, respectively, but there was never any dispute that
a GLO should be made.
727.2.At [13], Mr Justice Fraser observed that there had been a total of 10 interim
hearings before him in the 12 months prior to the first trial of the
substantive issues. This appears to have been intended as a criticism
levelled at both parties since he went on immediately to say that “/t/he legal
advisers for the parties regularly give the appearance of taking turns to
outdo their opponents in terms of lack of cooperation ... I made similar
comments in judgment No.1. These must have fallen on deaf ears, at least
for some of those involved in this case”. In relation to the strike out
application specifically, Mr Justice Fraser thought that “this counter-
productive approach lurks in the background to this application”, noting
that POL had first raised concerns about the scope of the Claimants’
evidence around a year earlier (which he felt showed “considerable, if not
Page 403 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
almost supernatural, foresight on [POL’s] part’): [14]. I make three points
in relation to this:
(i) First, it is correct that there were a number of interlocutory hearings
before Mr Justice Fraser in 2018, however most of them did not arise
out of applications made by either party. Save for the application to
strike out evidence and POL’s security for costs application which was
heard on 19 September 2018 (which concerned only the amount and
nature of the security to be provided, the Claimants having agreed to
security for costs in principle), all of the other interim hearings were
CMCs or costs budgeting hearings. On the whole, I do not recall these
hearings being excessive in number or especially fractious; rather, my
recollection is that they were a function of the staged approach to the
trial structure and disclosure process.
(ii) Second, as to the suggestion that the parties (including POL)
continued to approach the litigation in an uncooperative manner
following Mr Justice Fraser’s judgment [2017] EWHC 2844 (QB), I
believe that the approach that POL adopted was reasonable overall.
As I have explained above, after Judgment No.1 we took proactive
steps to ‘reset’ the relationship with Freeths (§390 and §392); we
asked David Cavender QC to review and critique our whole approach
to the litigation (see §391); we had constructive meetings with Freeths
to discuss the future approach to case management and disclosure
(see for example §538); and WBD advised POL to adopt a reasonable
stance on disclosure which sometimes went beyond what was strictly
Page 404 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
required by the Court’s Orders (see generally §§523-525 and §§537
ff above).
(iii
Third, there were points on which the parties did not agree, and the
parties’ differing approaches to the relevant factual matrix was a
prominent example of this. However, in my view there was nothing
inherently wrong with the parties sometimes having opposing views,
and it was fair and reasonable for POL not to simply concede its
position on the factual matrix. I do not agree with the characterisation
that POL displayed ‘supernatural’ foresight in alluding to this point in
late 2017, before the Claimants’ evidence for the CIT was served. As
I have explained above, it was evident from the pleadings and the
factual matrix document that the Claimants would seek to rely ona
wider range of factual evidence than POL’s legal team thought
permissible. In my view, it was appropriate to ventilate this point at an
early stage and not to hold it back until after service of the Claimants’
witness statements. Nor do I agree (if it were to be suggested) that
POL should have dropped the point and refrained from making the
strike out application following receipt of the Claimants’ evidence. As
I have set out above, whilst POL’s legal team appreciated that the
application was not without risk (and we advised POL of this), in our
view POL had little option but to make it given that the Claimants were
seeking to rely on a large volume of inadmissible material which POL
had not addressed (for the reasons I have explained) in its own
evidence. In light of earlier hearings before Mr Justice Fraser, and in
particular the 5 June 2018 CMC, my understanding was that he was
Page 405 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
sympathetic to POL’s position on the scope of the factual matrix, and
indeed that he had anticipated a strike out application as one possible
way to resolve the matter.
727.3.At [57], Mr Justice Fraser gave a general warning that “[ajn aggressive
and dismissive approach to such major Group Litigation (or indeed any
litigation) is entirely misplaced”. This was evidently directed at POL since
it was prefaced by the following remarks:
“Some passages of the Lead Claimants' evidence relate to the
circumstances in which their engagement with the defendant was
terminated ... The Lead Claimants complain that such
terminations were abrupt, came out of the blue, accused them of
falsifying accounts and made other statements that were not
factually accurate, and also that the defendant's approach (and
that of its solicitors) was generally heavy handed. I have read
some of this correspondence ... The tone of some of it is
undoubtedly aggressive and, literally, dismissive.”
I did not consider that the warning in [57] was fairly made. It was not based,
so far as I could see, on correspondence sent in the course of the group
litigation, but rather on the underlying factual matters in dispute in the Lead
Claimants’ cases. In my view, conscious efforts were made to ensure that
the tone of POL’s correspondence in the group litigation did not come
across as unduly heavy-handed or aggressive.
727.4.Mr Justice Fraser also referred to the fact that POL sought to have the
application dealt with at the beginning of the CIT, instead of at the hearing
that had been set down for 10 October 2018. I note that he described this
stance as ‘surprising’ ([9] of the Strike Out Judgment). Insofar as this was
intended as a criticism, I have explained the reasons why POL sought to
postpone the application to trial above at §722. In my view that stance was
Page 406 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
not unreasonable or irrational, notwithstanding that Mr Justice Fraser
preferred to case manage the application differently.
727.5.Finally, I address the suggestion that the strike out application was
motivated by a desire on POL’s part to ‘tailor’ the evidence, including in
order to limit the adverse publicity that might be generated by some of the
Lead Claimants’ evidence (cf. [32], [54]-[56] of the Strike out Judgment). I
do not believe that this suggestion was justified. I have set out the reasons
why the Counsel team and I recommended that POL should make (and
maintain) the strike out application above at §§720-722. As those
paragraphs and the advice cited therein make clear, our advice was not
based on the risk of adverse publicity to POL, but on the fact that the
Claimants were (in our view) seeking to prematurely lead evidence which
was irrelevant to the Common Issues. As part of that analysis, we
recognised that the evidence might have a potentially prejudicial effect on
the Judge’s mind, but this was a legitimate consideration and wholly
distinct from the wider risk of media interest in the Lead Claimants’
allegations. Further, we recognised that the evidence which was targeted
by the application would in due course become relevant to a future liability
trial; the application was therefore not about shutting that evidence out
altogether, but seeking to ensure that it was ventilated at the appropriate
time in line with the approach of dealing with the GLO issues sequentially
in stages.
728. With reference to Q94.4, I do not recall being challenged by representatives of
POL (including Paula Vennells) following the Strike Out Judgment. Although by
the time the application was heard the Counsel team and I had doubts as to how
Page 407 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
far it would succeed, the tone of the judgment came as surprise. Our expectation
was that even if the Judge refused the application (by applying a high threshold
for striking out evidence pre-trial), he was unlikely to regard it as being
unreasonable or unfounded. Indeed, our reading of matters was that the Judge
himself had anticipated that such an application might need to be made, and that
he broadly agreed with us on the question of what constituted admissible factual
matrix. Indeed, even after the application was refused, we understood that the
Judge had left it open as to what evidence he heard should be taken into account
in his decision on the Common Issues, and what evidence he should avoid
making findings about on the basis that this could trespass on a future liability
trial (cf. [53] of the Strike Out Judgment).’"? This is why we sought to maintain
POL's pre-existing position on admissibility and relevance at the CIT, rather than
adopting a different approach.
729. We did however modify our approach to the HIT evidence in light of the Strike
Out Judgment. At the time the judgment was handed down, the Claimants’
witness evidence for the HIT had been served around three weeks previously,
on 28 September 2018. As set out further below (§782), it included evidence from
individual SPMs which we considered was non-compliant with the Court's Order
following the 5 June 2018 CMC (which prohibited “Claimant-specific” as opposed
to “generic” evidence of fact).’2° The evidence which the Claimants served took
us by surprise and we had had no opportunity to properly investigate it, so we
were considering whether to apply to strike out the parts of it which were in
breach of the Order. In view of the Strike Out Judgment, we decided not to do so
719 Our understanding in this regard is evidenced by the Decision Paper in which we advised POL
not to appeal the Strike Out Judgment: WBON0001700.
720 POL00120352, paragraph 10.
Page 408 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
as the Judge had indicated that his intention was to address matters of evidence
at trial rather than through pre-trial procedural hearings. I also considered that a
second strike out application in short order would appear heavy-handed in light
of Mr Justice Fraser's comments. Instead, we sought to respond to the
Claimants’ evidence as best we could in the time available, principally through
Angela Van Den Bogerd’s evidence for the HIT.
(iii) Preparation of witness evidence (Q90.2, Q92 to Q93)
Overview of my role
730. My firm’s records demonstrate that it was Counsel who first drew up the plan for
POL’s generic witness evidence. This listed the issues that Counsel wanted to
be covered and divided them by reference to six different (generically described)
witnesses who might deal with them (for example, a “mid-level executive, familiar
with [the] nature and operation of PO[L’s] business’).’2' Victoria Brooks, a
Managing Associate in my team who supervised the day to day work on
preparing the CIT witness statements, then took the lead on assigning names to
the generic subjects in the evidence plan.’
731. Along with that generic evidence, the legal team planned to compile specific
evidence from POL staff in connection with the individual Lead Claimants where
it was both relevant to the Common Issues and possible to do so. As I have
outlined above, the firm view that was reached in consultation with Counsel was
that a great deal of the evidence proposed to be adduced by the Lead Claimants
721 WBONO001254.
722 WBON0000535 and WBON0000536.
Page 409 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
was inadmissible; as such, we proposed not to respond to that material in our
own witness evidence.
732. The primary responsibility for liaising with POL’s witnesses and preparing their
statements was divided between members of the team working on the GLO
matter: Ed Duffield, Helen Creech, Ivan Roots, Dave Panaech and Mandy
Robertson.’2° Victoria Brooks supervised this task and undertook a first level
review of the statements that they produced. From time to time, Victoria would
check in with me as regards the progress that the team was making with the
witness evidence.’24
733. There were particular challenges in collating the witness evidence for the CIT:
733.1.First, the general fallibility of memory, especially where witnesses were
being asked to remember specifics of matters that dated back many years.
For example, witnesses who gave evidence on how prospective SPMs
were interviewed (such as Elaine Ridge and Michael Haworth) undertook
dozens of interviews every year and so struggled to recall details of
specific interviews with individual SPMs. For that reason, at times the best
evidence those witnesses could give was as to their general practice when
conducting interviews.
733.2.Second, the sheer scale and complexity of POL's business and the
degree to which processes changed over time. This meant getting a clear
answer on the detail of how particular processes (e.g. appointment
processes) worked at a specific point in time was inherently challenging.
723 WBONO000540.
724 See for example WBON0000568 and WBONO0000609
Page 410 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
734. I was not involved in the initial stages of gathering the evidence and preparing
first drafts. I became much more involved after first drafts were completed,
conducting a detailed second level review of each statement personally and
coordinating further reviews by the Counsel team and Rodric Williams.’ All
statements were at minimum reviewed by Owain Draper and/or Gideon Cohen
who were junior Counsel for the CIT, as well as by Rodric Williams. At various
points, members of the team approached me with ad hoc queries in relation to
the preparation of witness evidence, for example as to the inclusion of particular
points.726
735. There were three witness statements in which I was more heavily involved than
others: Angela Van Den Bogerd’s, because she was to provide POL’s
overarching and principal evidence on how it operated; and two smaller witness
statements (those of Paul Williams and David Longbottom) which I picked up
towards the end of the drafting process. I cannot now specifically recall the
reasons for the latter two but I believe it had to do with capacity issues within the
team. I summarise my involvement in preparing each of those statements further
below.
736. I also dealt with overarching points of strategy, such as whether we should serve
particular statements, in respect of which I liaised with Counsel. On 14 August
2018, I contacted the Counsel team in connection with the draft statements of
Kendra Dickinson and Alison Bolsover, which dealt, respectively, with the
operation of the NBSC Helpline and with Transaction Corrections, Transaction
Acknowledgments, and POL’s debt recovery processes. I expressed the view
725 WBONO000569
726 WBON0000582
Page 411 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
that we did not require the level of detail that was included in these statements,
and, in any event, the material they contained was largely inadmissible because
it did not go to any of the Common Issues.’2’ I asked Counsel to briefly review
and advise on:
“1, Whether we need these statements at all?
2. If there are any sections/paragraphs in these statements that might
be useful? If there are short bits, we might be able to include these in
someone else’s statement.”
737. In response, Gideon Cohen stated that, in his view, they were almost entirely
inadmissible, and therefore “we should not serve either statement (or take any
sections for deployment in other statements).” From that point, we ceased work
on those statements. We also substantially cut down Helen Dickinson’s
statement to exclude material which, on reflection, we considered was not
relevant to the Common Issues.’”8 This reflected the approach recommended to
the Steering Group that I have discussed above (§§712-713), namely, that there
may be minor instances where evidence was on the borderline of inadmissibility,
but where this was judged to be too extensive we either cut it out or did not serve
the statement at all. Indeed, as noted above my recollection is that as the
evidence was prepared, the stance on not including inadmissible evidence
hardened.
Paul Williams and David Longbottom
738. My firm’s records show that it was Mandy Robertson, a solicitor in my team, who
was allocated the first draft of Paul Williams’ witness statement.’?° Paul Williams
727 WBONO000608
728 I comment further on this statement at §797.3 below.
729 WBONO001271.
Page 412 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
signed a finalised copy of his statement on 3 August 2018 without mine or
Counsel's approval out of an abundance of caution, because he was going away
at the time the witness statements were originally due (unless we secured the
extension of time which I was then seeking).”°° I see from my email records that
this was drawn to my attention at the time by Victoria Brooks.
739. We were subsequently granted an extension of time to 24 August, and Paul
Williams was to return before the revised deadline. As I recall, at this point there
was a Capacity issue; Victoria was busy working on other witness statements and
so I took over some of them. I reviewed Paul Williams’ evidence and concluded
that it did not address the “big questions” on Alan Bates’ case, namely “...
whether POL sent out the full contract terms with the Transfer Pack or not. If not,
how did [Mr] Bates get a copy of the full contract?”’*' From that point, I took over
producing a second draft of Paul Williams’ statement. I had a call with Paul
Williams on 14 August and amended his statement in light of that conversation.
I followed up later that day with a second draft of the statement and some further
questions.’*2 I completed two further drafts of Paul Williams’ statement, which
were informed by his own revisions and additional questions that I asked him by
email.”°
740. Once I had made those revisions, I sent Paul Williams’ statement to both Rodric
Williams and Angela Van Den Bogerd for their comments.’*4 Angela Van Den
730 WBON0000583
731 WBONO000606
732 WBONO0001281.
733 WBON0001289.
734 WBONO0001287.
Page 413 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Bogerd responded to confirm that Paul Williams’ statement accurately reflected
her understanding of how things operated.”°5
741. As to David Longbottom’s witness statement, my firm’s records indicate that
Victoria Brooks was tasked with preparing the first draft.’°° I reviewed that draft
in the usual way, and Victoria advanced it in accordance with my revisions.’°”
742. During the final stretch of the (re)drafting and review process in respect of David
Longbottom’s statement, on or around 21 August 2018, I took over management
of the statement from Victoria.’** As above, I believe this reflected a capacity
issue at the time. By this point, Counsel had fed into the statement by way of
amendments, though there remained some outstanding points to be checked
with David Longbottom. I sent an updated draft to him which he approved with a
few minor tweaks,’°° and dealt with Rodric Williams’ comments on the draft
statement.’4°
Angela Van Den Bogerd
743. Angela Van Den Bogerd’s witness statement comprised generic evidence that
offered an overview of how POL and SPMs operated. As I recall, and because
this was POL’s principal generic evidence, I was more heavily involved in its
preparation than the evidence of other witnesses. Further, by this point I had
spoken to Angela Van Den Bogerd dozens, if not hundreds, of times across a
significant period that spanned back to near the beginning of my involvement in
735 POL00041921.
736 WBON0001257.
737 WBONO000610, attaching a marked up version of the statement.
738 WBON0000612
739 WBON0001295 and WBON0000618.
740 WBONO000619.
Page 414 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
the Horizon-related matters. As such I believed that I had a good grasp of the
extent of her knowledge of POL’s operating practices.
744. My firm’s records show that Victoria Brooks and Mandy Robertson interviewed
Angela Van Den Bogerd on 15 January 2018 for the purpose of producing a proof
of evidence." A proof of evidence was produced, which Angela Van Den Bogerd
signed on 22 August 2018.’4? My firm’s records indicate that it was Ivan Roots,
working with input from Victoria Brooks, who then turned the proof of evidence
into a (partial) first draft witness statement.’”° It appears from my emails that I
had some discussions with Ivan around that time as to the structure of the
statement,” though I cannot now recall the content of any discussions Ivan and
I may have had, nor do I have any record of the same. It may be that I provided
this initial input on structure as lvan was new to the team working on the group
litigation and relatively unfamiliar with the overall case.
745. My firm's records indicate that I had carriage of Angela Van Den Bogerd’s witness
statement after that partial first draft was produced by Ivan Roots and Victoria
Brooks. On 24 July 2018, Victoria Brooks asked me to take over writing the
section of the statement dealing with implied terms.’4>
746. At the same time, Ivan was liaising with Angela Van Den Bogerd and other POL
employees, such as Kathryn Alexander, to further flesh out the factual content of
the draft statement. On 1 August 2018, Ivan sent me a copy of Angela Van Den
Bogerd’s statement “taken as far as [he] could take it’ for me to conduct a detailed
741 WBON0000517 and WBON0000515.
742 WBONO000546
743 WWBON0000556 and WBONO0000560.
744 WBONO000557.
745 WBONO000562
Page 415 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
review.’“° In completing that review I restructured the statement, added in new
sections (in particular, further background and context), and cut out some
material in order to strip out detail and omit any material that I considered to be
inadmissible.”4” Although the original proof of evidence had been prepared after
the setting down of the Common Issues for trial, it captured her evidence in a
broad sense but was later cut down in line with the strategy agreed with the
Counsel team on excising evidence which was not of direct relevance to the
Common Issues. Through this process, Angela Van Den Bogerd’s evidence was
substantially cut down and sharpened to focus on the matters that were known
by POL and SPMs at the time of contracting (or on knowledge that would have
been available to SPMs at that time).
747. I sent the revised version that I had prepared on to Angela Van Den Bogerd and
Rodric Williams for comment. At the same time, I also sent the statement onto
Gideon Cohen and Owain Draper.’“8 I also specifically asked David Cavender
QC to review the draft statement because of its centrality to POL’s evidence.’“°
Given both the length and importance of Angela’s statement, it went through a
significant number of iterations during the review process. During this time, Ivan
Roots and I continued to work in tandem on the developing draft.’°
748. I am specifically asked to consider one of those iterations of Angela Van Den
Bogerd’s statement, namely the version she sent to me on 20 August 2018
(POL00041956, attached to POL00041955). I am asked to consider two
748 WWBONO000580.
747 WBONO001286.
748 WWBONO000332.
748 POLO0363477
750 WWBONO000616
Page 416 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
comments on that version of her statement, which are highlighted in yellow and
thus are marked for Angela Van Den Bogerd’s attention:
748.1.First, a comment inserted at paragraph 67 of the statement, which reads:
“[Angela — are you comfortable with saying this]’ (cf. Q92.2). I think it is
likely that I wrote this comment, but I cannot specifically recall it. My firm’s
records suggest I did write it. It appears that this paragraph was inserted
along with the comment as part of my review of Ivan Roots’ first draft of
the statement. Looking at this comment now, I would consider that I likely
wrote it as I was unsure as to whether I had fairly captured Angela Van
Den Bogerd’s understanding of the position in paragraph 67. That
paragraph sketched in outline terms the communication between IT
systems in branch and the IT systems of POL’s clients. I sought to draw
her attention to my insertion so that she would review it and check it
accorded with her knowledge. Indeed, her comment in response to mine
indicates that she reflected on this insertion and sought external input (“/
will be if I have a bit more info on the detail and perhaps a couple of
examples of actual client requirements. Let me talk to a couple of people
before I commit’).
748.2.Second, a comment inserted at paragraph 115 of the statement, which
reads: “[Please feel free to amend below as appropriate]” (cf. Q92.3).
Again, I think that it is likely that I wrote this comment, though I cannot
specifically recall writing it. Reading it now, I would consider that I wrote it
as I thought that these paragraphs on Horizon fairly reflected Angela Van
Den Bogerd’s understanding, but I flagged the point as I wanted to draw
her attention to them. I can see through running a comparison of the
Page 417 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
previous version (dated 1 August 2018) and this version that although I
took some of the material from the Horizon section of that previous draft,
much of this was new drafting that I wished Angela Van Den Bogerd to
read carefully. By my edits, I sought to refocus the Horizon section of the
statement (as was my overarching approach) solely on information that
would have been known to a SPM at the time of their appointment. I
wanted Angela Van Den Bogerd to verify that what I had drafted in relation
to what a SPM would have known at this time was correct.
749. When Angela Van Den Bogerd’s statement was near the point of completion, I
had various further discussions with Counsel on the inclusion of specific points.
In particular:
749.1.On 22 August, I contacted Junior Counsel to discuss the section of her
statement which dealt with Horizon.”5' Previously, Junior Counsel had
deleted a number of paragraphs from the Horizon section, which dealt with
privileged user access and Angela Van Den Bogerd’s confidence in the
Horizon system.’* I was of the opinion that those sections should go back
into the draft statement. This was discussed on a call between Owain
Draper and me. I do not recall that call, but I can see by comparing the
version that was shared with Junior Counsel and the final version of the
statement that the final version was much reduced; that was in line with
the general approach I have outlined above.
749.2.0n 23 August 2018, David Cavender QC sent his and Gideon Cohen's
amendments to and comments on the draft statement. I was not
751 POL00363491.
752 POL00363453.
Page 418 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
comfortable with some of the deletions that had been suggested. In
particular, I raised the deletion of footnote 23 in paragraph 88, which
qualified the statement that “the Subpostmaster has complete control over
the branch accounts and transactions only enter the branch accounts with
the Subpostmaster’s knowledge...” in the following way:
“I put to one side here the Claimant’s allegations around Post
Office remotely editing branch accounts’ without a
Subpostmaster’s knowledge as these issues are the subject of
the Horizon Issues trial. I note however that these allegations,
which are put in a variety of different ways, are spurious or only
true in incredibly rare circumstances such that they are
inconsequential. I any event, they would not be in the mind of a
Subpostmaster when they joined the Post Office.”7°3
I explained my concerns about that deletion as follows:
“You've also deleted footnote 23 in para 88, regarding remote
access. If we do not cover remote access, then the entire
statement needs to changed, because the statement is premised
on the notion that SPMRs have control over their accounts. That
point is not 100% true, due to the remote access argument, but is
99.99% true so we need to explain the remote access concept
otherwise AVDB won't be telling the truth.
Given the sensitivities on this point, we need to cover it off.”">*
749.3.When I raised this point, Counsel agreed that the footnote ought to be
reinserted.’ A revised version of that footnote features in the final draft of
Angela Van Den Bogerd’s statement.
750. By this point in time, we had a near final version of Angela Van Den Bogerd’s
statement, subject to Rodric Williams’ comments and some outstanding minor
pieces of information that we required from her. In Rodric Williams’ email of 23
753 POL00363501.
754 WBONO0000625
755 WBONO0001311.
Page 419 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
August 2018 to which he attached a version of Angela Van Den Bogerd’s
statement that he had commented on (POL00041986), he made an overarching
point to the following effect (which I am asked to comment on by Q93):
“One overarching point on the witness statements: please make sure
you are giving the witnesses the “health warning” on signing a statement
of truth, i.e. they need to be confident that what they say in the statement
is true to the best of their knowledge and belief, and that they don’t
accept something just because it’s been through the lawyers” (emphasis
in original).
751. I do not know why Mr Williams made this overarching point, but looking at it now,
I consider it likely that he was simply being prudent. In any event, I can confirm
that I sent the following warning on signing a statement of truth to Angela Van
Den Bogerd when I attached her final statement for signature:
“A witness statement must have a "statement of truth". This is a
statement confirming that the person making it believes that the facts
stated in the document are true. The penalties for signing a statement of
truth without an honest belief in the truth of the facts being verified are
potentially severe. A person who makes a false statement in litigation in
an attempt to interfere with the course of justice will be in contempt of
court, which is punishable by a prison sentence of up to two years.”7°°
752. That was a warning (with some slight variations in wording) that WBD gave to
witnesses of fact in this litigation both before and after Rodric Williams sent the
email in POL00041986. This was part of our standard practice and was not
prompted by this email.’5”
753. By Q92.4 and Q92.5, I am asked to comment on a particular finding which Mr
Justice Fraser made at [544] of the Common Issues Judgment. As I have
756 POLO0363552
787 See by way of example: WBON0001274 (Paul Williams); WBON0001291 (Michael Webb);
WBONO0001296 (Tim Dance); WBON0001299 (Helen Dickinson); WBON0000623 (Michael
Shields); WBON0001301 (Michael Haworth); WBON0001307 (David Longbottom); see also the
information sheet provided to witnesses: POL00154271.
Page 420 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
highlighted above at §§697 fff (especially §§612-613), POL took the decision on
the advice of WBD and the Counsel team to confine its evidence only to those
matters which we regarded as admissible and relevant to the Common Issues;
as such, we did not propose to adduce witness evidence going to post-
contractual knowledge and events.’°* Angela Van Den Bogerd’s evidence was
prepared in line with this approach, and indeed we cut out large parts of her
original proof and earlier drafts of her statement for this reason.’®° That she was
not giving evidence on such matters in her witness statement was made clear at
paragraph 64, which read as follows:
“This is only a high-level overview, reflecting information which I believe
that new applicants for Subpostmaster could reasonably be expected to
know, or to have found out, before being appointed as a Subpostmaster.”
754. The criticism that Angela Van Den Bogerd had not offered the “whole story” in
her witness statement, or not included matters that were unfavourable to POL,
largely related to post-contractual matters and stemmed, in my view, from the
fact that questions were put to her in cross-examination about matters which
went beyond the factual matrix identified in paragraph 64 quoted above. In my
view it was not unreasonable to advise Angela Van Den Bogerd to confine her
witness statement in this way (and consequently it could not have been
unreasonable for her to accept this advice and limit her witness statement
accordingly), for the reasons I have explained above.
(iv) POL’s case on the effect of the ‘settle centrally’ button (Q90.3)
758 Save where permissible in respect of narrow exceptional areas (see above, §699, §712.3,
§714). This was the case, for example, in relation to paragraph 33 of the witness statement of
John Breeden, whose evidence responded to the Claimants’ "sham contract" points.
759 Save in one narrow respect which I comment on at §797.3 below.
Page 421 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
755. By way of background to this issue, much of the CIT (and indeed, the HIT) was
taken up with debate about how SPMs could dispute shortfalls. This flowed from
one of the Common Issues, which concerned whether the accounts produced at
the end of a trading period by SPMs (the Branch Trading Statement, or “BTS”)
could amount to a ‘stated account’ under ordinary agency principles. POL’s case
at the CIT on this matter never relied on the existence of a ‘dispute’ button in
Horizon. Indeed, POL accepted, as a fact, that there was no button in Horizon to
dispute shortfalls, but it maintained that this was irrelevant because there were
other means by which an SPM could raise a dispute about their accounts; in
particular, by calling the NBSC helpline. On POL's case, raising a dispute through
the helpline (or by any other means e.g. email, speaking to their line manager,
etc.) was properly to be regarded as part of the accounting process. We argued
that a BTS could therefore amount to a ‘stated account’ where no dispute was
raised over the BTS contemporaneously with its being lodged with POL. But,
whether any specific BTS lodged by any individual SPM was to be considered a
‘stated account’ depended on the facts of each case (namely, what the SPM said
and did at the time it was lodged). POL invited Mr Justice Fraser to go no further
than this during the CIT, and to leave questions around the submission and
disputing of accounts, and investigations into disputed shortfalls, to a later liability
trial.76°
756. However, the Claimants focused heavily on the fact that a shortfall could not be
disputed within Horizon itself, such that (they submitted) SPMs were forced to
lodge BTS's containing items with which they did not agree. On that basis the
Claimants argued that no BTS could ever amount to a ‘stated account’. They
760 Cf. paragraph 112 of POL’s written closing submissions.
Page 422 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Claimants made extensive submissions, supported by references to post-
contractual evidence and cross examination of POL witnesses, around the non-
existence of a ‘dispute’ button in Horizon. Ultimately this persuaded the Judge
to explore how SPMs could dispute Transaction Corrections and end of trading
period shortfalls during the CIT.
757. As the CIT advanced and Mr Justice Fraser's interest in this topic became
clearer, I suggested to the Counsel team by email on 13 November 2018 that we
pick this up in Angela Van Den Bogerd’s examination-in-chief.’' Gideon Cohen
responded that we had the material available to deal with this point in
submissions. I responded asking to see a copy of those proposed draft
submissions:
“Please can you send me the draft submissions (or an outline) on point
2. We really need to nail this point otherwise we risk our house of cards
coming down as [Patrick Green QC] will say if SPMs can't dispute losses,
then PO forces them to put inaccurate things in their accounts, that
justifies false accounting, it also means that they cannot be held liable
for what their accounts say (because those accounts are plainly
inaccurate because they can't dispute mistakes), thus the normal rules
of agency cannot apply to SPMs and clauses 12.12 / 4.1 should be
construed as requiring PO to prove every loss in every account.””®
758. Owain Draper responded outlining the shape of those submissions (which he
proposed to write the following day) as follows:
“The key points are (1) the manuals in all relevant periods make clear
that amounts settled centrally can be disputed and (2) the Lead
Claimants did in fact settle centrally and dispute. Ultimately, therefore,
the rules were clear and the Lead Claimants acted consistently with
having known the rules. Any suggestion now that they did not know that
they could dispute is self-serving and can be rejected. It is also worth
noting that many (perhaps even all) the LCs in fact admit to having
761 WBONO000635
762 WBONO0000636
Page 423 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
known that they could dispute amounts that they settled centrally (which
is unsurprising given that they in fact did it).’"®>
759. On 6 December 2018, which was the fifteenth and final day of the CIT, Mr Justice
Fraser asked the parties to devise an agreed flowchart that laid out the steps
involved in the Transaction Correction and BTS processes. This was one of many
tasks Mr Justice Fraser set the parties in the course of the trial. I do not recall
how that document was prepared. My firm’s records suggest it was Amy Prime
and Dave Panaech in my team who coordinated with Junior Counsel to prepare
the relevant flowcharts.’ On 14 December 2018, following the end of the trial,
Amy Prime sent our proposed flowcharts to Freeths (with me in copy).7° My
firm’s email records suggest that there was some disagreement between the
parties on the wording of the ‘settle centrally’ box in the Transaction Correction
flowchart. The version of the wording that Amy had sent to Freeths read as
follows:
“Settle Centrally”
(1) A credit or debit entry for value of the TC is made in the SPM’s branch
account; and
(2) A corresponding debit or credit is made in the SPM’s customer
account with Post Office.
SPM may phone NBSC to lodge a dispute. If this is not done, Post Office
will contact the SPM to discuss payment of any shortfall.”
760. Freeths’ responded on 17 December with various comments on the Transaction
Correction flowchart, one of which was that they had “[cJhanged the wording in
763 WWBONO0000637.
764 WBON0001379.
785 WWBONO000642.
766 WBONO0000643.
Page 424 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
relation to “Settle Centrally” to more neutrally reflect the position’.’*’ Their
proposed wording was as follows (I show the changes in red):
“Settle Centrally”
(1) A credit or debit entry for value of the TC is made in the SPM’s branch
account; and
(2) A corresponding debit or credit is made in the SPM’s customer
account with Post Office. If a debit, this will be treated as a debt by Post
Office
SPM may phone NBSC to lodge a dispute. /fthisis-netdenePoest Office
il tact the SPM to oii tof. hortfall.”
Pay eA
761. Owain Draper responded that this change was not remotely neutral and indeed,
it was contrary to our pleaded case (I discuss the relevant paragraph in the
Generic Defence above at §§500 ff).”®8 Dave Panaech offered up the suggestion
that we accept Freeth’s wording, with the following minor change to point (2)
(shown in yellow highlighting):
(2) A corresponding debit or credit is made in the SPM’s customer
account with Post Office. If a debit, this will be treated as a debt by Post
_ ; rs
Office Unless” NV. dge a di
762. Counsel approved Dave’s suggested wording. He reverted to Freeths’ on those
terms. Freeths’ reverted as follows (1 have shown Freeths’ proposed addition in
red):
“We believe the position on the evidence, was that an amount settled
centrally but disputed, results in collection of the debt being suspended,
but factually a debit is considered by Post Office to be a debt. We
propose the wording below, which we believe fairly reflects this.
“(2) A corresponding debit or credit is made in the SPM's customer
account with Post Office. If a debit, this will be treated as a debt by Post
787 WBON0001393.
768 WWBONO000645.
768 WBONO000646 .
Page 425 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Office unless the SPM contacts NBSC to lodge a dispute, which should
suspend collection of the debt until the dispute is resolved.”
Otherise we are willing to agree your position for the sake of ensuring
agreed documents can be filed as was clearly the Judge’s preference.”
763. Again, Counsel regarded this wording as contrary to POL’s pleading, and Owain
Draper suggested that the words “of the debt’ be removed to sidestep the
semantic point.’”° Freeths’ version incorporating Owain Draper's suggested
deletion was the wording in the final version of the Transaction Correction
flowchart that was filed with the Court.
764. As already noted, I discussed the drafting of the relevant paragraph of the
Generic Defence dealing with this point (paragraph 43(3)) above at §§500 ff.
There is a tension between the Generic Defence and the flowchart on this point.
The Generic Defence pleaded that an amount settled centrally but disputed via
the NBSC helpline was not a debt at all and was not treated as such by POL,
whereas by the end of the CIT (as the flowchart demonstrates) the position had
changed slightly, viz. it was strictly speaking a debt, albeit one that POL would
not enforce until any dispute was resolved. During the CIT, I saw this as a largely
semantic point as to what was properly classified as a ‘debt’. As set out above, I
do not recall this fine distinction being raised by anyone (including POL) at the
time of the pleading; the main point being that if a shortfall was disputed, POL
would not, in practical terms, chase for payment of it (or at least, that was what
POL's operating procedure provided; although POL sometimes failed to follow
this and did chase disputed sums, and this was heavily emphasised by the
Claimants in the CIT).
770 WBONO0001396.
Page 426 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(v) Approach to cross-examination of the Claimants, including allegations of
dishonesty (Q90.4)
765. The proper approach to cross-examination of the Lead Claimants fell within
Counsel’s remit and not my own. I cannot recall whether or not I received David
Cavender QC’s cross-examination notes in advance, nor does my firm have any
record of my doing so. The approach that Counsel generally adopted to cross-
examination was, as I understood it, a continuation of POL’s wider stance on the
proper scope of evidence for the CIT (which I have discussed above at §§697 ff).
Following the Strike Out Judgment, POL (and in particular Counsel) had a tricky
line to tread in terms of maintaining a clear and strict line on inadmissibility when
it came to cross-examination, as it was maintained at trial that evidence of events
which took place after the Claimants contracted with POL was generally
irrelevant to the Common Issues, yet parties are entitled to challenge evidence
that they dispute. Thus, had POL failed altogether to cross-examine evidence led
by the Claimants on post-contractual matters which we considered to be
irrelevant and inadmissible, and the Judge found otherwise, POL would have lost
an opportunity to put its case on that evidence.
766. We had raised with POL the risk that this type of cross-examination could make
its case appear confused — this was one of the reasons for seeking to strike out
the Claimants’ inadmissible evidence before the CIT began as that risk would
then be avoided.’”' After the strike out application was rejected, David Cavender
QC still advised that the best approach was for him to put what he described as
771 POL00256731.
Page 427 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
‘blocking’ questions in his cross-examination of the Lead Claimants to mark
POL's disagreement with the offending parts of their evidence.’’2
767. In the course of cross-examination of the Lead Claimants, David Cavender QC
also put documents to some of them that went to their credibility as witnesses,
and in the cases where there was evidence of false accounting (i.e. dishonesty)
those points were also put. That strategic decision was properly within Counsel's
remit, who had significant experience of how cross-examination was conducted,
including the reasons for cross-examining on credit. There is a relevant email
chain on this topic between Amy Prime and David Cavender QC (with me in
copy), which concerns discussions as to the proposed contents of the Lead
Claimants bundle for the CIT. Amy states in her email of 18 September 2018 that
WBD has “inserted the documents which have been disclosed and are dated
prior to the contract being entered/the C entering the branch”.’’> David Cavender
QC responded as follows:
“Whilst I can see the logic of this — I think there are some documents
which are post contractual that we might want to put to witnesses for
reasons of prejudice.
By way of example: cases such as Sabir and Abdula : the documents
such as audit reports revealing their deficits — how they were caused —
what they said at the time etc: these will be useful in asking questions
which will assist on the optics of the case and go to the credibility of the
witness.
I will try and identify as I go through the cases those documents in this
category that it might be useful to include to assist you in this process.”””4
772 This approach was explained to the Steering Group in a Decision Paper advising on the
prospects of appealing the Strike Out Judgment: WBONO0001700.
773 WWBONO001328.
774 WBONO0001331.
Page 428 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
768. In closing submissions, Counsel were careful not to rely on matters which we
considered were outside the scope of the relevant and admissible factual matrix.
Further, Counsel stressed that Mr Justice Fraser should avoid making any
findings (including as to credit) that risked trespassing on future trials, for
example because such findings would go to issues of breach and/or causation.
Accordingly, POL made very limited submissions on credibility and did not rely at
all on evidence of false accounting in relation to credibility (see further below,
§§ 981-982).
P. PREPARATION FOR THE HORIZON ISSUES TRIAL
769. In this section I deal with the preparation of POL’s case for the Horizon Issues
Trial, which took place over various dates between March and July 2019. This
section therefore addresses Q95 to Q102 of the Request, save that I have
already dealt with Q95.1 and Q99 above, in Section N on Disclosure (at §§666-
670 and §§672-694, respectively).
770. In particular, this section: (i) gives an overview of the preparation of the evidence,
including POL’s case on bugs, errors, defects and remote access (Q95.2 to
Q95.3); (ii) addresses how and to what extent POL drew on Fujitsu in preparation
for the HIT, including the decision not to call Gareth Jenkins as a witness (Q95.4,
Q98); (iii) answer’s the Inquiry’s questions about specific aspects of the witness
evidence, namely an email sent by Gareth Jenkins on 16 November 2018
containing information that was then reflected in Angela Van Den Bogerd’s
witness statement for the HIT (“AVDB2”) (Q97), my involvement in the
preparation of Steve Parker’s evidence, and inconsistencies and inaccuracies in
Page 429 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
the witness evidence on remote access (Q101); (iv) my involvement in the
preparation of Dr Worden’s expert evidence (Q96, Q102); and the merits advice
POL was given in the lead-up to the HIT (Q100).
(i) Preparation of the evidence — overview (Q95.2 to Q95.3)
Overview of the preparation of POL’s evidence
771. In the course of preparing this statement, I have refreshed my memory as to how
these matters were handled with reference in particular to two Steering Group
papers which were produced about the preparation of the evidence for the HIT,
one dated 20 September 2018””° and the other dated 12 October 2018.76
772. Due to the technical nature of the issues in dispute at the HIT, most of the Horizon
Issues were a matter for expert evidence. Some of the issues however also
required factual evidence, and certain other matters were purely factual. At the
22 February 2018 CMC (see §321 above), Mr Justice Fraser identified that the
March 2019 trial window should be used to address issues about Horizon, and
indicated that it should be based predominantly on expert evidence.
“I wanted the parties to agree or propose an isolated number of issues on
the pleadings related to Horizon that would involve expert evidence but
not evidence of individual cases ... I was obviously not sufficiently clear so
I am going to make it clear now. My intention is in March to resolve the
Horizon Issues that observe the following three criteria. Issues regarding
the Horizon system that arise on the pleadings, that is the first one;
second, that can be resolved on expert evidence, third, do not require
evidence of fact or if they do require the barest evidence of fact.”"””
775 POL00257368.
776 POLO0257886.
77\VBONO0001337.
Page 430 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
773. One of the difficulties that POL faced from the beginning of the litigation was that
the Claimants did not detail the nature of their allegations about Horizon in their
pleadings, and this made it difficult for POL to know the case that it had to meet.
By early 2018, the extent of the case put forward by the Claimants in relation to
Horizon was the sparsely pleaded details in the Amended GPOC and Generic
Reply.’”8 At some stage I expected the Claimants to put forward examples of
when they had encountered unusual behaviour by Horizon or transactions /
accounting entries that they could not explain and therefore suspected were the
result of a bug. From there, I expected the approach would be that the experts
would look at each of those examples and to give an opinion on whether there
had been an error in Horizon. Indeed, I asked Mr Coyne about adopting this
approach at the meeting between the parties' lawyers and experts in April 2018,
but it became apparent that he had not received this information from any of the
Claimants and was not in favour of approaching the Horizon Issues in this way —
although he was also not able to clearly articulate the approach he wished to
take (see paragraph §614 above).
774. Accordingly, at the CMC on 5 June 2018 (see §322 above), the lack of clarity in
the Claimants' allegations was discussed with Mr Justice Fraser. His view was
that the HIT was to be predominantly a trial of expert evidence and he did not
expect to be determining the details of individual Claimants’ cases. He therefore
ordered that each party file "witness statements of any witness of fact whose
generic evidence (in distinction to Claimant-specific evidence) they wish to rely
upon for the purposes of determining the Horizon Issues" (emphasis added).’7°
778 See §377.3 above about this information not being included in the SOls.
779 POL00120352, paragraph 10.
Page 431 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
775. The Court had also made other orders to help draw out the Claimants’ case on
Horizon. By 18 July 201876 (subsequently extended to 17 August 2018)’®' the
Claimants were to produce an outline document "setting out the nature of their
allegations in relation to the Horizon Issues" This document was produced as
directed on 17 August 2018 but shed little light on the Claimants’ allegations
about Horizon, and gave no hint that the Claimants were planning to adduce
evidence from individual SPMs (being the approach their expert had opposed at
the meeting in April 2018 and which had then been prohibited by the Order
flowing from the 5 June 2018 CMC, cited above).’®2
776. In view of these matters, the Court also ordered a sequential exchange of factual
evidence and expert reports beginning from September 2018.’® After the first
round of factual evidence by both sides (“Round 1”, exchanged on 28 September
2018), the Claimants’ first expert report (Coyne 1) was filed in October 2018, then
POL had the opportunity to file responsive witness statements (“Round 2”,
served on 16 November 2018) and after that, POL’s expert report (Worden 1)
was filed on 7 December 2018.
777. As a result of this sequencing, WBD advised POL to keep its first round of
evidence limited to background information about Horizon and to then provide
more evidence in the second round once we had sight of the Claimants’ export
report and the Claimants’ own witness statements. Thus, in the Steering Group
Paper dated 20 September 2018 (which I prepared), I explained that:’64
780 POL00117925 at paragraph 15.
781 WBONO001269.
782 POL00358213.
783 POL00117925.
784 POL00257368.
Page 432 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
"The Claimants have served a provisional outline document setting out
the nature of their allegations in relation to the Horizon issues, but this is
lacking in detail and the Claimants have not articulated their case on
Horizon adequately. We have asked them to clarify this case and they
have refused. This lack of clarity makes it difficult for us to know what
evidence will be needed to respond to the Claimants’ arguments until we
see their expert report.
LJ]
The lack of detail about what the Claimants will allege in their expert
report combined with the opportunity to submit further evidence at the
end of October, lead us to believe that Post Office should serve minimal
evidence in the first round. This evidence would cover basic
background information only and will provide Robert Worden with the
evidence that he requires to produce his report."
778. I was also aware of POL’s unavoidable reliance on Fujitsu for its factual evidence
about the Horizon system. There was no one at POL who could cover the Horizon
system, the known bugs, remote access, or Horizon controls. However, POL's
dependency on Fujitsu’s evidence came with the risk that the accuracy of that
evidence could not be fully tested. The Steering Group paper explained:
“Fujitsu have obvious reasons for ensuring that their evidence is
accurate, and we know that their own legal team is closely monitoring
the position. Nevertheless, there is a residual risk that Fujitsu could
(inadvertently) put forward inaccurate evidence. Given that this evidence
is of a technical nature (about Horizon) or regarding Fujitsu's internal
practices, it is difficult for us to validate this evidence other than asking
probing questions in the usual course of taking evidence. To be clear, we
have no reason to believe the evidence will be inaccurate, but given that
it is coming from a third party it is important that this risk is understood.”
779. In view of these matters, WBD strongly recommended that POL only submit
minimal evidence in Round 1. All of the other options were so fraught with risk
that they were not worth any real consideration:
"The alternatives to [the] above minimal approach are:
Page 433 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Submit no evidence. We believe that this would lead to criticism from the
Judge who is clearly expecting Post Office to submit some evidence.
Submit full evidence. This would require a degree of guesswork on the
topics to be covered which is not attractive at all.”
780. I discussed this with Rodric Williams, and he agreed with this assessment.
Therefore, rather than seek a formal decision from the Steering Group in respect
of this course of action, the Steering Group paper updated them that this was the
course of action that POL had instructed us to take.
781. In line with this approach, POL's Round 1 evidence for the HIT, served on 28
September 2018, was intentionally limited. For example, Torstein Godeseth’s first
statement (“Godeseth 1”) described the Horizon system at a high-level only
(including the ways in which Fujitsu could alter transaction data and the controls
around audit data).
782. As it turned out, this was a sensible approach to have taken. The Claimants'
witness evidence, served on 28 September 2018, included evidence from six
SPMs about problems they had encountered in branches which they suspected
were caused by Horizon (as well as an ex-Fujitsu employee, Richard Roll). In
my view, the evidence of the individual Claimants largely breached the Court's
Order prohibiting "Claimant-specific evidence". No prior warning was given by
the Claimants that they would be calling these witnesses, and accordingly no
disclosure orders had been made in relation to them and no investigations had
been undertaken by WBD into their cases at this point.’® Further, although the
witness statements raised important issues to be investigated — especially in
785 Indeed, this led us to consider whether POL ought to apply to strike out the Claimants’ evidence
to the extent that it breached the Court's Order. However, in view of Mr Justice Fraser's Strike
Out Judgment (discussed above at §723 and §§726-729), we decided not to do so and to
respond to the Claimants’ evidence as best we could in the time available.
Page 434 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
relation to the evidence of Richard Roll — it was still not clear how the Claimants
intended to formulate their case in relation to Horizon at trial.
783. Indeed, it was not until Coyne 1 was served on 16 October 201878 that (i) it
became clear that the Claimants were going to focus their case heavily on the
contents of the KEL on the footing that it showed evidence of Horizon not working
properly, and (ii) they began to identify the KELs they believed were relevant. Mr
Coyne did not say that there was some fundamental design flaw in Horizon or
that there was some significant bug that undermined all (or even a majority of) of
transaction / accounting entries. Rather, it appeared to me that the Claimants'
and Mr Coyne's approach was to say that the KEL showed there were a large
number of smaller bugs, although he was ambiguous about what this meant for
Horizon’s robustness overall. He concluded (para 3.7) that:
"Whilst the present-day version of Horizon, supported by manual human
support may now be considered as relatively robust in the spectrum of
computer systems used in businesses today it has undergone major
modifications in its history. It is likely that in 1999 when it was first
commissioned, and in 2010 when it was significantly upgraded (to
Horizon Online), it was less robust.
784. He also said that "Horizon’s relative robustness does not mean that [sic] is
thereby extremely unlikely to be the cause of shortfalls" but did not go on to give
an opinion that Horizon was likely to be the cause shortfalls in any given set of
branch accounts. I inferred from the absence of an opinion on this critical point
786 POL00258234.
Page 435 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
that Mr Coyne may have believed that Horizon was unlikely (albeit not extremely
unlikely) to be the cause of shortfalls.”8”
785. In my mind, this was the point when we first had a true sense of the Claimants'
case on Horizon and when the Horizon Issues Trial started to take shape. Whilst
Coyne 1 raised numerous, troubling new points that needed investigation, it did
not assert the conclusion that Horizon was not robust and so I recall that my
general impression on first reading Coyne 1 was still that POL had a credible
case that Horizon worked reliably.
786. For POL’s Round 2 evidence (due on 16 November 2018), we were largely
responding to the Claimants' evidence and expert report. We had an evidence
plan that we regularly updated which set out key points of evidence which we
were seeking for each witness to give.”8° This was a significant exercise to
complete in the 6 weeks, given (i) the shape of the Claimants' case only started
to surface in their recently served evidence, (ii) Richard Roll raised serious
allegations about precise technical aspects for Horizon that had not previously
been put by the Claimants, (iii) Mr Coyne's expert report cited dozens of KELs
that needed to be analysed, and (iv) to respond to the evidence from the
individual SPMs required an investigation of their cases from a standing start.
This work was being run in parallel with supporting Dr Worden's preparation of
his expert report, and preparing for and conducting the CIT.
787. The purpose of Stephen Parker's first witness statement (“Parker 1”), served in
Round 2, included responding to the evidence of Richard Roll and to give
787 Mr Coyne later appeared to accept this under cross examination at the HIT, where he said "Yes,
the vast majority of all the transactions that flow through the Horizon system will work
successfully": HIT Transcript, {Day 14/58:17} to {Day 14/59:15}.
788 WBON0000632.
Page 436 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
evidence on the then-known bugs (the Receipts and Payments Mismatch bug,
the Suspense Account bug, the Callendar Square bug, and the Dalmellington
bug). The purpose of Torstein Godeseth’s second statement (“Godeseth 2”),
served in this round, was to respond to the evidence of Charles McLachlan (the
defence expert in the prosecution of Seema Misra), and to also give evidence on
the known bugs. Stephen Parker and Torstein Godeseth were witnesses of fact;
their evidence was not to offer an opinion on the efficacy or accuracy of Horizon,
this being a matter solely for the expert witnesses.
788. With that context, I turn to the Inquiry’s Q95.2 to Q95.3, concerning the
preparation of POL’s witness statements and its case on bugs, errors, defects
and remote access.
Preparation of the witness evidence (Q95.2,
789. The legal team’s preparation of the evidence for the HIT occurred whilst we were
also preparing for and conducting the CIT, so we had to split the team. Tony
Robinson QC led the Horizon Issues team, whilst David Cavender QC focused
on the Common Issues. Owain Draper and Gideon Cohen (Junior Counsel)
stayed predominantly focused on the Common Issues, until that trial finished and
then Owain assisted with the Horizon Issues. Recognising the need for more
resource for the Horizon Issues team, in May 2018 we instructed Simon
Henderson, who had deep experience of IT litigation, and in July 2018 we added
a further specialist IT Junior Counsel in Rebecca Keating. At WBD, Jonny
Gribben led on the preparation of the HIT evidence with input and advice from
Counsel.’®° Different members of the WBD team were assigned different
788 WBON0000629; WBONO0000630.
Page 437 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
witnesses. Lucy Bremner was assigned Torstein Godeseth and Jonny was
assigned Stephen Parker.’°°
790. My own involvement in preparation of evidence varied depending on the round
of evidence:
790.1.Round 1: the exchange of the first round of witness evidence on 28
September 2018 was around a month before the CIT. To the best of my
recollection, I reviewed all of the witness statements that POL served, but
I was nonetheless heavily focused on preparation for the CIT.
790.2.Round 2: following exchange of the Round 1 statements on 28
September and then Coyne 1 in October 2018, POL’s responsive
evidence was served on 16 November 2018. The turnaround time for this
evidence was therefore short — originally just over four weeks and then
extended to seven weeks from exchange of the Round 1 evidence, and
only four weeks from service of Coyne 1. My supervision of the
preparation of evidence in this round was much more limited as I was in
the middle of the CIT. Counsel took a significant role in reviewing the
statements and generally advising on the approach to the evidence
(including in some instances speaking directly to witnesses).’°' I cannot
recall what exactly I did or did not look at, but I may not have reviewed
everything; in some cases, I may have only seen early drafts. As I now
recall from having reviewed the documents to refresh my memory in order
to prepare this statement, a lot of the witness statements were being
finalised at the last minute, with changes to the evidence often being made
780 WWBONO000627.
791 For example: WBON0001694.
Page 438 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
on the day of service (see e.g. §§830-835 and §856 below). I would not
have had the time to have reviewed every single item of correspondence
with redrafts going back and forward between WBD and a witness.
790.3.Round 3: a limited number of further statements were prepared and
served in January and February 2019. I reviewed the draft statements, but
this was heavily driven by Counsel, who by this time were under brief for
the HIT.
791. I recall reviewing several different versions of the witness statements at different
points in time, but cannot say that I reviewed them all in detail primarily due to
being focused on the CIT, especially the Round 2 statements served on 16
November 2018 when the CIT was underway. My best recollection is that I was
not involved in interviewing Stephen Parker or Torstein Godeseth. Likewise, I
was copied into lots of emails with Fujitsu about the evidence, but I would not
have had time to review and comprehend such dense technical information. I do
not recall whether I reviewed the email exchanges between Jonny Gribben and
Stephen Parker or Lucy Bremner and Torstein Godeseth whilst those witness’
evidence was being prepared, but generally speaking, unless an email was
directly addressed to me or something was specifically drawn to my attention, at
this point in time, I was not able to review every email which I was copied into. I
was receiving dozens of emails per day at this time (often more on days where
there were key deadlines such as for filing of evidence), and I was in the midst
of the CIT.
Overview of the preparation of POL’s case on bugs, errors and defects (Q95.3)
792. Below I give an overview of POL’s case on bugs. I set out the legal team’s
assessment of the merits of POL’s case (in particular, on the central question of
Page 439 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Horizon’s robustness) as we headed into the HIT below at §§899 ff; these
answers therefore also summarise my response to Q6 to Q9 of the Request, as
they relate to the period heading into the HIT.
793. Three bugs had been acknowledged by POL from the outset of the litigation: the
Suspense Account bug, the Receipts and Payments Mismatch bug, and the
Callendar Square bug. In particular, POL admitted the existence of these bugs in
its LOR dated 28 July 2016 (having previously disclosed them to Second Sight
during the Spot Review process). Similarly, POL’s pleaded case admitted that
Horizon was not a perfect system and did not seek to contend that it never had
any bugs.
794. By Autumn 2018, the experts had received disclosure from the KEL and Peak
databases, and lots of other disclosure from Fujitsu in particular. The review of
this material for bugs was largely left to them as the experts. In Coyne 1, Mr
Coyne cited lots of KELs which he described as being of "significant interest" but
his analysis of those KELs was not clear about whether some or all of these were
evidence of bugs.’%? In response, Fujitsu reviewed 58 of the KELs cited by Mr
Coyne, and a further 48 selected by Dr Worden, and produced two tables
(appended to Parker 1) setting out its views on whether they were evidence of
bugs and, if so, whether they had an impact on branch accounts. In many cases,
they concluded there was no bug or it had no impact on branch accounts. This
analysis was undertaken separately from the IT experts’ own reviews of KELs
and searches for bugs.
792 POL00258234, paragraph 5.114.
Page 440 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
795. Worden 1, dated 7 December 2018,’ involved, in my view, a much more in-
depth review of KELs and Peaks than did Coyne 1. Dr Worden dismissed most
of the KELs identified by Mr Coyne as either not being evidence of a bug at all or
not having an impact on branch accounts (or that the impact would have been
quickly corrected by safeguards already built into Horizon).’** He did however
identify 11 bugs in Horizon that had an impact on branch accounts (including the
already acknowledged bugs).’°> Dr Worden's view was that this was a very small
number of bugs relative to the scale of Horizon. His conclusion was that "Horizon
has been a very robust system, compared to other major systems I have worked
on in sectors such as banking, retail, telecoms, government and healthcare"’*>
and that "the robustness of Horizon made it extremely unlikely to be the cause
of shortfalls in branches".’” My view at this time was that Dr Worden's report
was more persuasive than Coyne 1, and that Mr Coyne's report was at worst
(from POL’s perspective) equivocal about whether Horizon was robust. Having
read both reports, I continued to believe that POL had a credible case that
Horizon was robust.
796. Mr Coyne's supplemental report dated 1 February 2019, Coyne 2, identified 22
bugs.”°8 Coyne 2 was longer than Coyne 1, and was not in my view a
‘supplemental’ report. Instead, it advanced a much more detailed and structured
criticism on Horizon, now centrally focussed on identifying not only KELs and
Peaks of "significant interest’, but ones that evidenced a bug in Horizon that
783 POLOOI11481.
794 Ibid, paragraph 733.
795 Ibid, paragraph 742.
796 Ibid, paragraph 49.
797 Ibid, paragraph 68.
798 POL00262929.
Page 441 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
impacted branch accounts. Mr Coyne’s conclusion was that "Horizon is less
robust than as originally expressed in my first report".”°° However, he did not say
that Horizon was not robust or indeed provide an opinion on how reliable or
accurate Horizon was. My reading of Coyne 2 was that Mr Coyne's views were
still ambiguous. Saying that Horizon is "/ess robust" than he originally thought,
did not in my mind mean that Mr Coyne was saying that Horizon was unreliable.
I still therefore believed that POL had the better of the argument on Horizon.
797. The scale of the new bugs Mr Coyne purported to find did however cause me
concern, as this was substantially more than I had known about at the start of
the group litigation or was expecting might surface during the litigation. I believed
that even if all 22 truly were bugs, this was still a very small number compared to
the scale of Horizon based on the analysis done by Dr Worden. But the fact that
they had come out late in the day, and were not pro-actively raised by POL from
the outset, left an impression that POL and Fujitsu were either not being
forthcoming about material facts or did not have a good understanding of their
own system. Either way, this was damaging to the credibility of POL's case.
798. On 25 February 2019, the experts agreed a Second Joint Statement that
produced a codified list of 29 potential bugs found by both experts.®° By this
point (just weeks before the start of the HIT), it was too late to submit more
evidence on bugs, so we took the approach of preparing notes of each alleged
bug that could then be used in cross-examination or closing submissions. These
notes were prepared by WBD by reviewing the relevant KELs and Peaks,
gathering Fujitsu’s (and sometimes Atos’) input, preparing what became known
799 Ibid, paragraph 1.2.
800 POL00266866.
Page 442 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
as “bug notes”, and then having Fujitsu check each note. When ready, the notes
were then passed to Counsel. The notes formed the foundation to the Appendix
to POL's written closing submissions on all of the identified bugs. Dr Worden also
analysed the bugs identified by Mr Coyne from his own perspective.
799. Fujitsu were also asked to comment on the bugs which had been identified.
There was no particular point person — WBD’s requests went to Fujitsu and were
picked up by whoever at Fujitsu was best placed to assist. However, WBD kept
Fujitsu’s input separate from Dr Worden in order to ensure that both experts
(Worden and Coyne) had equal access to information. We did not, therefore,
permit Dr Worden and Fujitsu to discuss the bugs as that would have given Dr
Worden more information than Mr Coyne.
800. This analysis began before the HIT commenced and was completed during the
course of the HIT. As a result of our (Dr Worden, Fujitsu and WBD’s) analysis of
these 29 bugs, POL's closing submissions concluded that (i) eight were not bugs
at all, (ii) three had no impact on branch accounts, (iii) nine had a transient impact
and (iv) nine had the potential to cause lasting impact.®" This analysis reassured
me that the scale and impact of the 29 potential bugs was not as substantial as
presented by Mr Coyne in his supplemental report.
801. On 1 March 2019, shortly before the HIT started, the experts produced their Third
Joint Statement. Their central conclusion was that they agreed that "From our
experience of other computer systems, Horizon is relatively robust".8°2 In my
mind, this was a significant agreement of the experts in that Mr Coyne was, it
appeared to me, accepting that Horizon was a generally good IT system. I was
801 See POL00278807, F12 to F18.
802 POL00026918.
Page 443 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
very surprised that Mr Coyne had agreed to this given that his supplemental
report had worked hard to identify as many bugs as possible. I interpreted this
as meaning that Mr Coyne must have believed that the scale of the bugs he had
found was sufficiently small not to undermine the overall reliability of the system.
This reinforced my view that POL had a reasonable case that the Court would
find Horizon to be robust.
Overview of the preparation of POL’s case on remote access (Q95.3)
802. Certain forms of remote access had been acknowledged by POL from the start
of the group litigation. In particular, POL’s LOR acknowledged that there were
ways in which Fujitsu could influence branch accounts remotely (see §474
above). As set out above, this was approved by Deloitte. Similarly, POL’s Generic
Defence (approved by both Deloitte and Fujitsu, see §§493-494 above) was that:
802.1.Fujitsu could not remotely log on to a Horizon terminal in a branch so as
to conduct transactions.
802.2.Fujitsu had the ability to inject a new transaction into a branch’s account
(called a Balancing Transaction in Horizon Online), which is logged and
extremely rare, but that functionality did not extend to editing or deleting
data in those accounts.
802.3.Further, POL pleaded that there was a small number of Fujitsu specialists
who had privileged user access rights which in theory enabled them to
amend or delete transaction data for a branch, but to do so and to conceal
the steps taken thereafter was so difficult, and it involved such complex
steps, that it was extremely unlikely to have occurred.
Page 444 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
803. POL’s case on remote access had been developed based on the findings of
Deloitte, who had investigated remote access functionality as part of its work on
Project Zebra and was then asked in 2016 to 2017 to conduct further
investigations about the robustness of Horizon, including remote access
functionality, for the purpose of advising POL in its defence in the litigation
(Project Bramble). I describe those investigations above at §§202 ff, §§466-472,
§§483-490 and §§510-520.
804. From the date on which POL’s defence was drafted up to the preparation of the
evidence for the HIT, I was not aware or made aware of any substantial new
findings about remote access functionality.
805. I set out further below (at §§841 ff) the evidence of Torstein Godeseth and
Stephen Parker on the different means of remote access, which included the
revelation of the ability of Fujitsu personnel to remotely access a branch terminal
(counter) in Old Horizon so as to inject a new transaction.
806. The parties’ experts also gave evidence on remote access functionality. They
both agreed that remote access was possible, but there was a divergence of
views about how often it had happened. I did not direct Dr Worden on how to
address this issue, but did comment on drafts of his reports, including the remote
access sections. I deal further with the Inquiry’s Q96, concerning the nature and
extent of my involvement in the preparation of Dr Wordens’ reports, below at
§§862 ff.
(ii) The extent of Fujitsu and Gareth Jenkins’ assistance in preparing POL’s case,
and the decision not to call Gareth Jenkins (Q95.4, Q98)
Page 445 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Fujitsu’s involvement (Q95.4,
807. With reference to Q95.4, POL prepared its case on Horizon based on the advice
of its legal team and the information provided to it by Fujitsu which was, in turn,
tested by Deloitte, and later, Dr Worden and Mr Coyne. Neither Fujitsu nor Gareth
Jenkins helped “prepare” the case, but they did provide factual information we
needed to understand to put the case together.
808. Fujitsu’s role was limited to being a source of knowledge about how the Horizon
system worked. In practice, that meant that Fujitsu personnel were involved in
the following ways:
808.1.Fujitsu personnel engaged directly with Deloitte to respond to queries and
questions about Horizon.
808.2.Fujitsu answered questions put to it by WBD on various points as and
when they arose. This was typically by email.
808.3.Fujitsu provided WBD and Deloitte with access to its technical documents.
808.4.Stephen Parker, Torstein Godeseth, Andrew Dunks and William Membery
provided witness statements.
808.5.Fujitsu reviewed the “bug notes” referred to above which were prepared
by WBD, and provided comments.
808.6.Fujitsu met with Mr Coyne and Dr Worden to brief them on Horizon and to
facilitate inspections of the KEL and Peak systems.
Page 446 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
809. On 4 February 2019, WBD had an all-day meeting with Fujitsu and Counsel in
order to discuss a number of technical issues in advance of the trial.®°> The
purpose of this session was not to prepare the case and there were no strategy
discussions at this meeting. It was limited to assisting Counsel to better
understand the technical materials, particularly in respect of KELs and Peaks.
Decision not to call Gareth Jenkins (Q98)
810. As to the decision not to call Gareth Jenkins as a witness (cf. Q98), POL’s legal
team, including the Counsel team, were unanimously agreed that he should not
be called.
811. I had become aware in 2013 that Gareth Jenkins had given incorrect evidence
in past criminal proceedings against SPMs, in that he had asserted that Horizon
Online was ‘bug-free’ despite being aware of two bugs which had affected that
system (namely, the Receipts and Payments Mismatch bug and the Suspense
Account bug): see above, §§86-88. The Clarke Advice, prepared by Simon
Clarke of Cartwright King in 2013, had considered the evidence he gave in a
number of past prosecutions and concluded:
“— Dr. Jenkins failed to disclose material known to him but which
undermines his expert opinion. This failure is in plain breach of his duty
as an expert witness.
- Accordingly Dr. Jenkins credibility as an expert witness is fatally
undermined; he should not be asked to provide expert evidence in any
current or future prosecution.
- Similarly, in those current and on-going cases where Dr. Jenkins has
provided an expert witness statement, he should not be called upon to
803 We had also held a conference between Counsel and Fujitsu in advance of preparing the
Generic Defence, see above at §481.4(iii).
Page 447 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
give that evidence, Rather, we should seek a different, independent
expert to fulfil that role.’®°*
812. Quite apart from his awareness of the two bugs in question, I had never
understood how Gareth Jenkins could have given evidence to the effect that an
IT system was completely error-free (i.e. perfect). As a matter of basic common
sense there was always at least a risk of there being a bug in any large IT system,
and indeed this reality was reflected in the approach that POL had taken during
both the Mediation Scheme and the group litigation, where its position was that
Horizon was not perfect and had suffered bugs, but was nevertheless generally
reliable.
813. It was clear to me that as a result of his past conduct in relation to criminal
prosecutions, Gareth Jenkins’ credibility as a witness was fatally compromised
such that he could not be stood up as a witness in the group litigation. Counsel
agreed. Tony Robinson QC had been provided with a copy of the Clarke Advice
with his original instructions on 1 June 2016.8°5 On 7 September 2018, I sent this
advice again to Tony Robinson QC and also to Simon Henderson as we were
then beginning to prepare POL’s evidence of fact for the HIT.8°° On 10 September
2018, Rodric Williams, Counsel, and I held a conference with Cartwright King to
enable them to share their views on Gareth Jenkins with Counsel.®” I recall that
the meeting lasted for about a couple of hours and that Cartwright King were of
the strong opinion that Gareth Jenkins should not be called as a witness as his
credibility was damaged. Tony Robinson QC concluded at that conference that
804 WBON0001723.
805 WWBON0001005 and WBONO0001011.
806 WBONO001315.
807 This was preceded by a background briefing email that I sent to Rodric Williams: POLO00042010.
Page 448 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Gareth Jenkins should not be called and advised Rodric Williams of this, who
agreed with this approach.
814. Other than this conference, the question of whether Gareth Jenkins should be
called as a witness was discussed with Counsel on other occasions, and every
time it was discussed, the reasons for not doing so were validated.
815. The problem that this presented for POL was that Gareth Jenkins was routinely
held out by Fujitsu as being their foremost expert on Horizon. He had (as I
understood it) to a large extent designed the system, and there were few people
if any at Fujitsu with the same level of knowledge about Horizon’s earliest days.
Sometimes Fujitsu would say that Gareth Jenkins was the best, or even only,
person who could answer some of the most intricate points about Horizon,
especially Legacy Horizon. By the time of the group litigation, he had retired from
Fujitsu but I believe he was under some form of consultancy agreement whereby
Fujitsu could seek his input on an ad hoc basis, and he had provided memos or
picked up enquiries by email. I do not believe that I met or spoke with him at any
point during the group litigation, and his input was generally provided in written
form.
816. Since Gareth Jenkins’ retirement from Fujitsu, Torstein Godeseth was generally
put forward by Fujitsu as the primary person who could answer technical
questions about Horizon; for example, it was Torstein Godeseth that attended the
conference with Counsel on 22 June 2017 prior to service of the Generic Defence
(see §481.4(iii) above). My impression was that Torstein Godeseth was very
knowledgeable about the system, having worked for Fujitsu on the Horizon
system for many years.
Page 449 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
817. Consequently, at the time of the conference with Cartwright King on 10
September 2018, I believed that Torstein Godeseth would be able to cover much
of the same ground as Gareth Jenkins, but I was aware that there would be some
blind-spots which could be problematic, especially around the very early years of
Horizon where Torstein Godeseth’s knowledge did not run as deep as Gareth
Jenkins’. Later in 2018, when we were preparing POL's witness evidence, I
learned that Stephen Parker had also been at Fujitsu for a very long period of
time and was also able to answer questions about Horizon.
818. In the event, as we drilled deeper into the evidence, particularly for Round 2 in
November 2018, Torstein Godeseth and Stephen Parker began to defer more to
Gareth Jenkins. This meant that more points than expected had to be based on
Gareth Jenkins ' input. This had not been my expectation given that Torstein
Godeseth and Stephen Parker had been held out as having deep knowledge of
Horizon and had both been working with Horizon for over a decade.
819. This posed a problem for POL because we wanted to limit the reliance on Gareth
Jenkins as much as possible, for the same reasons that POL had decided not to
call him as a witness. Tony Robinson QC captured the point in an email dated 12
November 2018, where he explained:
“We all know the reasons why we have decided not to have Jenkins as
a witness. They are also reasons for not having him as a source of
evidence — i.e. as a source of information for our witnesses and/or as a
person providing analyses on which our witnesses will rely. Where he is
acting as a source the Claimants will know this and they will waste no
time in arguing (1) the fact that we have not called such a natural witness
demonstrates that he is not a reliable witness, (2) we recognise this fact
and want to protect him from any cross examination and (3) if he is not
a reliable witness, he can’t be a reliable source of evidence, either and
(4) as the claimants are being prevented from cross examining him the
information he provides to other witnesses is even less reliable than a
Page 450 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
witness statement from him would be. This argument will undermine the
evidential value of any witness statements that are based on information
that Jenkins has provided. ”®°8
820. Tony Robinson QC advised that, in addition to the decision not to call Gareth
Jenkins as a witness, “we should limit Jenkins’ involvement as a source of
evidence as much as possible, essentially to those areas where there is no
alternative source of information” ®°° The problem was, as it became clear, there
were certain points on which POL needed Gareth Jenkin’s input, particularly on
issues relating to Old Horizon. In response to Tony Robinson QC’s email, Jonny
Gribben stated:
“We note the risk involved with using Gareth as a witness and we are
limiting Gareth's involvement as much as possible, but he is Fujitsu's go-
to person for many of our questions. If Torstein or Steve covered the
bugs they would still need to speak to Gareth (Torstein less so).”*'°
821. Efforts were made to minimise reliance on Gareth Jenkins by ensuring that the
matters which Stephen Parker and Torstein Godeseth put in evidence were, as
much as possible, based on first-hand knowledge. It was not however possible
to fashion a perfect solution in this regard, however, since even Stephen Parker
and Torstein Godeseth between them did not have the full range of knowledge
required for the purposes of the Horizon Issues. Thus, when POL’s Round 2
evidence was served on 16 November 2018, Tony Robinson QC observed after
having reviewed Torstein Godeseth’s second witness statement:
“TG’s statement is peppered with references to Gareth Jenkins as a
source of evidence. Given that such a large proportion of TG’s evidence
is not within his knowledge but is merely passing on his understanding
of what Jenkins has written or told him, we can expect some
uncomfortable questions as to why we have not called the organ grinder
808 WWBONO000342.
808 WWBONO0000342.
810 WBON0000341
Page 451 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
and why the monkey’s evidence should be given any weight at all. I
wonder whether the claimants will have the courage to demand that we
call Jenkins as a witness. As I didn't previously appreciate how every
road seems to lead back to Jenkins, I also wonder whether, depending
on how much of a fuss they make before the trial, should we review our
prior decision not to call him.’®"'
822. By a further email dated 7 December 2018, Tony Robinson QC circulated a list
of issues for the legal team to consider. Regarding witness evidence, he noted
that there were Horizon Issues on which (i) no witness evidence had been given
or (ii) inadequate evidence had been given and cited a “good example” of (ii) as
being “the evidence in which witnesses simply repeats what Gareth Jenkins has
told the relevant witnesses”. He identified that, “the fact that we have obviously
decided not to call Jenkins is going to be a problem for us at trial’.5'? In reality,
both the fact that Fujitsu had offered POL no witness evidence on some issues
and inadequate witness evidence on others stemmed from the fact that we were
not in a position to call the person at Fujitsu who knew the most about Horizon.
As a result, the evidence of the witnesses who were called was incomplete,
and/or reliant on information obtained from a source whom we had decided (for
good reason) not to call.
823. In relation to those areas where it was necessary to obtain information from
Gareth Jenkins, i.e. because the Fujitsu witnesses and other Fujitsu personnel
did not have the requisite knowledge, I believe we acted reasonably in seeking
Gareth Jenkins’ input. Despite his damaged credibility as a witness, I still
believed that he was a useful source of information. He was undoubtedly very
knowledgeable about Horizon, and was able to give detailed and authoritative
811 WBON0000189
812 WBONO001721.
Page 452 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
answers about the technical workings of the system. To have not sought his input
would have left holes in POL's evidence and would have been to disregard a key
repository of relevant knowledge. That of course needed to be balanced against
Tony Robinson QC's (and my own) concerns about the reliability of any
information sourced from Gareth Jenkins and the fact that his input was
effectively being voiced through Torstein Godeseth and Stephen Parker. We
therefore sought to rely on him only where his input was genuinely required, and
the nature of his input was to provide factual explanations about how technical
processes within Horizon worked (as opposed to any views he might hold about
the reliability or robustness of the system). Looking back, I accept now (and
recognised then) that this was an unsatisfactory state of affairs. The unfortunate
reality was that there was no alternative witness with the requisite depth of
technical knowledge within Fujitsu (and there was certainly none at POL). In
these highly unusual circumstances the approach taken — whilst clearly very far
from ideal — was ultimately a reasonable one in my view.
(iii) Gareth Jenkins’ email to Jonny Gribben of 16 November 2018 and
preparation of Angela Van Den Bogerd’s second witness statement (Q97)
824. Q97 of the Request asks me to comment on the following statement that Gareth
Jenkins made in an email to Jonny Gribben on 16 November 2018 at 12:28
(POL00111371):
“There is a further scenario. On Old Horizon if SSC were to insert a
transaction at the counter (which although possible, was very rare), then
this would have been associated with the User Id of whoever was logged
on at that counter. If nobody was logged on then the User Id would be
missing. Such transactions should be clearly identified in the audit trail
as having been inserted by SSC.”
Page 453 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
825. I was not copied into this email. It set out Gareth Jenkins’ answers to a series of
questions asked by Jonny Gribben, and Gareth Jenkins’ comment about
injections via the counter was one of these answers. I was only copied into Jonny
Gribben’s response timed a few minutes later (12:34), in which he stated, “in
relation to the section that I’ve highlighted in yellow below, can you explain which
reconciliation process should have picked the issue up? Is it possible that the
process would have picked this up in due course, but Mrs Burke was proactive?”
The text Jonny had highlighted in yellow did not include the paragraph quoted
above. The highlighted text related to a separate issue about the source of a
shortfall suffered by one of the Claimants giving evidence at the HIT, Mrs Burke.
826. There was then a further exchange between Jonny and Gareth Jenkins (again
solely focused on the issue relating to Mrs Burke), to which I replied, at 14:19:
“The bit I don’t understand is why Mr Burke was correct to handover the £150 to
the customer if the Recovery Receipt doesn’t show the transaction for £150?”
Gareth Jenkins answered that point to which I replied, “Thanks Gareth — I get it
now!”
827. I have no recollection of this email chain, but what I am sure I must have done is
read Jonny's email into which I was copied, been prompted to consider the
highlighted text which drew my attention (and which I would have surmised was
the reason for Jonny copying me into the exchange) and responded with my
comments in relation to that highlighted text. As I would have understood it at the
time, and as I understand it now upon reviewing POL00111371 for the purposes
of preparing this statement, this was the only outstanding issue arising out of
Jonny’s original questions to Gareth Jenkins. Certainly, I would have understood
it to be the only issue on which Jonny sought my input. In other words, I would
Page 454 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
have thought that everything else was in hand, and would have had no cause to
examine the rest of the exchange.
828. Nor would it have been practicable for me to descend into the detail of the rest
of the email exchange. For context, the email in POL00111371 was sent on 16
November 2018, which was (i) the day POL’s Round 2 evidence was due to be
served, and (ii) was during the CIT. It was a Friday, so I was not in court, but I
was extremely busy, and nearly all of my focus was on the ongoing trial. At that
time, I would often spend several hours each day on the phone to Rodric Williams
or Counsel, and I did not review every email to which I was copied in full unless
something specific was drawn to my attention. Indeed, a search of my email
records for 16 November 2018 indicates that I received around 120 emails that
day alone, and so I could not realistically have read through the detail of
everything that I received that day. More generally though, my role was to lead
the work the firm was doing and take overall responsibility for its delivery. I would
monitor progress, acknowledge in broad terms the work that was being
undertaken, and address issues where my input was specifically sought by my
team — but beyond this, but I would not and could not review everything I was
copied into in detail.
829. In the event, as §826 above shows, Gareth Jenkins provided a satisfactory
answer to my query about the issue relating to Mrs Burke’s recovery receipt, so
as far as I was concerned, there were no more outstanding issues arising out of
the email chain that required my attention. I therefore do not believe I read Gareth
Jenkins’ statement elsewhere in his email of 12:28 about injections via the
counter, and I certainly have no memory of doing so. For the same reasons, I do
not believe I discussed this aspect with Jonny Gribben.
Page 455 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
830. After having reviewed the email in POL00111371 and my firm's records for the
purposes of preparing this statement, I now understand that Jonny Gribben’s
original email was posing questions for Gareth Jenkins in order to enable POL to
finalise its Round 2 witness statements for the HIT, which were due later that day.
Specifically, I now understand (although this was not apparent from the face of
the chain in POL00111371) that one of Jonny's questions was relevant to Angela
Van Den Bogerd’s witness statement for the HIT, AVDB2, in that she had been
asked to set out in her statement the circumstances in which a transaction might
not be associated with an SPM’s User ID. Jonny mentioned two scenarios to
Gareth Jenkins (namely, where SPMs shared their User ID with others physically
in the branch, and situations where a second user logged on after an earlier
user’s session had disconnected), and asked him to identify any others. It was
this that prompted Gareth Jenkins’ response quoted above.
831. Earlier that morning I had reviewed a draft of AVDB2 which, in relation this point,
referred to the two scenarios mentioned by Jonny and then said (at paragraph
18.3):29
“lam not aware of any other reason the user ID could be affected. [Note:
we are awaiting clarification from Fujitsu as to whether there are any
other reasons the user ID could be affected].”
832. Katie Simmonds subsequently sent me and Jonny Gribben an email stating that
we had now received information from Gareth Jenkins and that she was going to
update the draft of AVDB2 accordingly.®"4 My firm’s records do not indicate that !
responded to that email.
813 WBON0000196; WBONO000197.
814 WBON0000287.
Page 456 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
833. Subsequently, Katie sent a revised draft of AVDB2 which contained the following
paragraph (now paragraph 18.4) in place of paragraph 18.3 quoted above:
“There is a further very rare scenario, in relation to Legacy Horizon only,
involving the insertion of a transaction at the counter by the SSC. In this
instance Horizon would associate the transaction with the user ID of the
individual logged on at that counter. If nobody was logged on at the time
the transaction was inserted, then the user ID would be missing. These
transactions would be clearly identifiable in the audit trail as having been
inserted by SSC.”815
834. Again, my firm’s records do not indicate that I commented on that draft or
responded to Katie’s email. I see that Katie then sent the statement to Angela
Van Den Bogerd who indicated that she had reviewed the draft and was
comfortable with the changes made, and signed the statement shortly
afterwards.®16
835. I cannot recall when I reviewed the version of AVDB2 which contained this
paragraph, though I have no recollection of doing so (and I have not identified
any emails to indicate that I did) before it was signed and served on 16 November
2018. If I did read it, I certainly did not appreciate the significance of paragraph
18.4 at the time. Nor did I appreciate that it drew on Gareth Jenkins’ email in
POL00111371 (which, as I have said, I do not believe I would read on 16
November 2018 in any event). Had I have been aware of this at the time, I would
have reminded my team of the need to check that Angela Van Den Bogerd was
either personally aware of the information in paragraph 18.4, and if not, that this
should be made clear in her statement.
815 WBON0000285; WBON0000286.
816 WBONO000195.
Page 457 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
836. Much later, on 12 February 2019, Simon Henderson informed me that it had
recently been suggested to him (it is not clear by whom) that Angela Van Den
Bogerd did not have personal knowledge of the matters contained in paragraph
18.4.5'7 This then gave rise to enquiries about whether these matters were in fact
within her knowledge or not. She subsequently confirmed that they were,®'® but
in hindsight, consideration of the attribution of the statement in paragraph 18.4
should have been fully explored in the first place given the similarity of that
paragraph to the email sent by Gareth Jenkins contained in POL00111371.
(iv) My involvement in the preparation of Stephen Parker’s evidence (Q101.1)
837. I was involved in arranging for Stephen Parker to give evidence but I had limited
involvement in the preparation of his evidence itself. I reviewed and commented
on an early draft of Stephen Parker's first statement (Parker 1),°"9 and I was later
involved in advising on certain decisions that needed to be made in respect of
how to address the inaccuracies in Parker 1 and his second witness statement
which was served in January 2019 as part of Round 3 (“Parker 2”). To the best
of my recollection, that was the extent of my involvement in Stephen Parker's
evidence.
838. Stephen Parker became a potential witness after Richard Roll’s first witness
statement was served in September 2018. On 12 October 2018, Jonny Gribben
met with Stephen Parker and Pete Newsome (also of Fujitsu) to discuss the
817 WBONO0000292.
818 WBONO0001432.
818 WBONO0000288.
Page 458 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
allegations made by Richard Roll. Jonny Gribben then turned the points
discussed in that meeting into a potential statement for Stephen Parker.®°
839. Stephen Parker was reluctant to give evidence. In an email to Jonny Gribben on
16 October 2018, he stated, “/ have seen this process previously where a
colleague signed such a statement which resulted in a very stressful court
appearance. I am happy to continue supporting the process and refining the
information but I will not be signing a witness statement, we need to find another
way to use this information.”®2' As a result, on 30 October 2018, I prepared a
briefing to Paula Vennells, asking whether she could speak to Fujitsu in order to
persuade Stephen Parker to give evidence or find another witness to help. The
key points of that briefing note stated:
“Mr Roll's evidence is that Horizon is defective and that is highly
damaging to Post Office's case.
He is a former Fujitsu employee, talking about Fujitsu internal operations
and therefore only Fujitsu can put up a witness to counter these
allegations.
The ideal person is Stephen Parker of Fujitsu, but he is refusing to be a
witness. Can you help persuade him to support us? Or can you find
another witness to help?
If we can't find a witness, Post Office will have to summons Steve to give
evidence. That will be damaging for PO's legal case, embarrassing for
Fujitsu and even more stressful for Steve.
This needs to be fixed urgently as there is an imminent Court deadline
for Post Office's evidence. Can you come back to me on this by
tomorrow?
LJ
820 WBON0000192; WBON0000193.
821 WBON0000194.
Page 459 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Stephen Parker of Fujitsu was identified as a good witness. Steve was
Mr Roll's team leader in 2003 and is now head of Fujitsu support services
centre. He can therefore give evidence of:
Mr Roll's position in 2003 and why, from his relatively junior position, he
is not a credible witness.
The true scope of Fujitsu's role in supporting Horizon, both in 2003 (when
Mr Roll was there) and over the next 15 years.
Steve has been (and continues to be) very helpful in preparing
responses to Mr Roll's allegations. From our interactions with Steve we
believe he will come across as a measured, knowledgeable and
articulate witness.
[uJ
Stephen Parker is refusing to be a witness. He has a friend who had a
very difficult experience of giving evidence in Court. We also understand
that Steve does not consider himself a natural public speak [sic] (he has
a slight stammer). He therefore simply does not want to do it for personal
reasons.
LJ]
Our preference therefore is for Steve to agree to be a witness, even if he
does so under protest. Ultimately however this is outside of Post Office's
control, and we need Fujitsu senior management to step in to make this
happen”.®22
840. Then, on 31 October 2018, Pete Newsome called me to say that Stephen Parker
had agreed to be a witness, and we could therefore “stand down Paula” 83
(v) Changes in Fujitsu’s position on remote access (Q101.2 to Q101.4)
The nature of the changes (Q101.2)
841. In the course of describing my developing knowledge about remote access at
e.g. §§ 202-232, §§271-274, and §§466-472 above, I explain the various changes
822 POL00258674.
823 WBONO0000284.
Page 460 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
in Fujitsu's position over the years. Necessarily, my knowledge developed in line
with these changes in position. The point I wish to add to this evidence which I
believe will be of assistance to the Inquiry relates to the specific changes in
position in Fujitsu’s evidence for the HIT. In summary, up to preparing the second
round of POL’s evidence, I had believed, and POL had maintained in its Generic
Defence (the wording of which was approved by Deloitte and Fujitsu, see §§493-
494 above), that it was not possible for Fujitsu to remotely log on to a Horizon
terminal in a branch in a way that would allow them to conduct transactions.
What changed (in simple terms, the technical detail I have now forgotten) was
that Fujitsu belatedly acknowledged that it could inject transactions via the
counter (i.e. the branch terminal) and, in effect, this meant that Fujitsu could
remotely conduct transactions in effectively the way that had previously been
denied. Further, conducting transactions in this way could (in some
circumstances) record those transactions against an SPM's or assistant's user
ID, potentially making it look like they had conducted the transactions — a point
which POL had previously denied.
842. Godeseth 1, which was finalised on 27 September 2018 as part of the Round 1
evidence, purported to comprehensively set out all the methods of remote access
— however, whilst he explained the injection of transactions at the
correspondence server, he did not mention the possibility of injection of
transactions via the counter. In relation to server injections, he also stated:
“[l]n legacy Horizon, any transactions injected by SSC would have used
the computer server address as the counter position which would be a
number than 32, so it would be clear that a transaction had been injected
in this way’.
Page 461 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
843. Parker 1, meanwhile, which was dated 16 November 2018 and served as part of
the Round 2 evidence also did not mention injections via the counter.
844. The above positions were then corrected in the Round 3 evidence. Paragraph
27 of Parker 2 dated 4 February 2019 corrected the omission in Parker 1 to
mention the “rare” process whereby Fujitsu Support (also known as “SSC”) would
insert transactions via a counter:
“In paragraph 20 of Roll 2, Mr Roll describes a process by which
transactions could be inserted via individual branch counters by using
the correspondence server to piggy back through the gateway. He has
not previously made this point clear. Now that he has, following a
discussion with colleagues who performed such actions I can confirm
that this was possible. I did not mention it in my first witness statement
because, when faced with a less clear account in Mr Roll's first
statement, my recollection was that if it was necessary for the SSC to
inject a transaction data into a branch's accounts, it would have been
injected into the correspondence server (injecting via the server was the
default option which was followed in the vast majority of cases).
LJ]
Transactions injected into a counter would appear on the transaction
logs available on Horizon as if it had been carried out by the user that
was logged into the counter at the time (if nobody was logged on, the
User ID would be missing). However, when injecting such a transaction,
the SSC user would ensure that it was clearly identified in the audit trail
as having been inserted by SSC. Examples of such identification I am
aware of are the use of a SSC user as the Clerk ID and/or details of the
incident number as an additional property.”824
845. In his third witness statement dated 28 February 2019 (“Godeseth 3”), Torstein
Godeseth made a correction to similar effect: “! have read Parker 2 I am now
aware that it was also possible for SSC to insert transactions with a counter
824 POL00266514.
Page 462 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
position with a number less than 32”. He explained that he had not been aware
of this capability at the time of making his earlier statements. °°
846. Parker 2 contained another correction to Parker 1. At paragraph 32 Stephen
Parker stated, in relation to the injections by the SSC:
“At paragraphs 21 and 22 Mr Roll states that both he and the "SSC team
generally had the ability to inject data" and that "there was no limit on the
type of transaction that we could insert". At paragraph 20.2 of my first
statement I said that "some" members of the team could do this,
but this was badly stated. Everyone in the SSC team had the ability
to inject data. My intention was to express the fact that only limited
numbers of SSC technicians ever needed to inject financial data”.®6
847. His third witness statement dated 28 February 2019 (“Parker 3”) made certain
further clarifications or corrections to Parker 1 and Parker 2. The relevant
sections stated:
“In paragraph 19 of my first witness statement, I stated that it was not
possible in Legacy Horizon to edit or delete data that had been
committed to the message store. I have been asked to clarify this
statement.
... In some circumstances ... it would be necessary for the SSC to delete
the message store file (and hence all the transaction data it held) to allow
Riposte to replicate a fulland complete copy of that transaction data from
another source. This process does not allow any partial deletion this is
an all or nothing operation; it is a similar process to recovering all your
data from a backup.
! do not consider the removal of incomplete or corrupted storage files, to
allow the built in facilities of the system to recover from alternative
copies, to be the deletion of transaction data” (at [20]-[22]) (added
emphasis).”
And:
“In paragraph 35 of my second witness statement I stated that, in theory,
someone could have used a transaction injection in Legacy Horizon to
825 WBONO0001218.
826 POL00266514.
Page 463 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
carry out a transaction such as a GIRO bank transfer or a utility bill
payment.
In a footnote I explained that GIRO bank transactions are automated
payment (AP) transactions, like utility bill payments, and that other bank
transactions go through a different path, as that was my understanding
at the time.
Having discussed the issue further with colleagues, I now understand
that GIRO bank transactions were EPOSS transactions (like all other
manual bank transactions) rather than AP transactions. The distinction
is that copies of AP transactions are sent to the Post Office client,
whereas with EPOSS transactions they are not” (added emphasis).”
848. The crucial change in evidence, and the most significant inconsistency in the
witness evidence, related to the injection of transactions via the counter and the
question of whether this process was always visible to the SPMR:
848.1.Parker 1, paragraphs 21 and 22, stated:
“21.1 Any transaction that was inserted would immediately cause
a discrepancy to arise in the branch's accounts. For example, if a
transaction were to be inserted which stated that £1,000 of
stamps had been bought by a customer who paid cash, that would
immediately cause a reduction in stock levels of stamps in that
branch and the branch would have £1,000 less in cash than
Horizon expected it to have.
21.2 In other words, although a transaction could be inserted, it
would immediately become apparent that this had been done
and ultimately it would not benefit any member of staff to
behave in this way.
22. It is correct that the "remote access" described above could
have been carried out without the permission of a Subpostmaster.
However, any additional transactions inserted remotely
would be identifiable as such from the transaction logs that
are available to Subpostmasters from Horizon.”
848.2. Parker 2 meanwhile stated:
“transactions could be inserted via individual branch counters by
using the correspondence server to piggyback through the
Page 464 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
gateway... Transactions injected into a counter would appear on
the transaction logs available on Horizon as if it had been
carried out by the user that was logged into the counter at
the time (if nobody was logged on, the User ID would be
missing).”
849. Knowing what I know now, Godeseth 1 was incorrect in that it failed to mention
injections via the counter when purporting to give an exhaustive explanation of
all the methods of remote access. At the time that Godeseth 1 was drafted, I did
not know about this second method of injection. To the best of my belief and
based on the documents that I have reviewed for the purposes of drafting this
statement, I do not believe that Jonny Gribben (who assisted Torstein Godeseth
with the preparation of his statement) knew either. Paragraph 25 of Godeseth 3
states that Torstein Godeseth did not know about this method of injection at the
time of drafting his first witness statement, and I have no reason to believe that
this statement is incorrect.
850. Parker 1, meanwhile, was incomplete in that it did not say that there was a
method of injecting directly via the counter which might make it appear as though
the injection had been conducted by the user logged on at the time.
851. The important point to stress is that the significance of this point (namely, whether
a transaction was injected directly into the counter or from a correspondence
server and or whether it bore a counter ID, the SPM’s User ID, or no ID) was not
understood by me, nor I believe by my team, until it became apparent in or
around February 2019.
852. It is also important to note that the existence of counter injections was set out at
paragraph 18.4 of AVDB2 to which I have referred above (§833) and therefore
was in evidence. I have quoted that paragraph above but to reiterate, it stated:
Page 465 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
“There is a further very rare scenario, in relation to Legacy Horizon only,
involving the insertion of a transaction at the counter by the SSC. In this
stance Horizon would associate the transaction with the user ID of the
individual logged on at that counter. If nobody was logged on at the time
the transaction was inserted, then the user ID would be missing. These
transactions would be clearly identifiable in the audit trail as having been
inserted by SSC.
In relation to transactions inserted by the SSC from the data centre in
Horizon, again, this would have no associated user ID, however, will be
clearly identified in the audit trail and will also be visible in branch reports,
including the transaction log, as having originated from the data centre
as opposed to a counter.”
How the inaccuracies in the evidence came about and who bears responsibility
(Q101.2, Q101.3):
853. Undoubtedly, the fact that counter injections were addressed in AVBD2, omitted
in Parker 1, and omitted in Godeseth 1 (in circumstances where the maker of
that statement was purporting to exhaustively list all the methods of remote
access), was regrettable. These were, however, points of deep technical detail
regarding an IT system. With evidence of this nature, I believe that a lawyer is
heavily reliant on a witness' knowledge and recollection of the details to ensure
that their statement is accurate.
854. As to Godeseth 1, as I have explained above, he explained that was unaware of
the existence of the counter method of injection and I have no reason to doubt
that was the case.
855. As to Parker 1, Steve Parker’s failure to mention counter injections was ultimately
his omission, which (as I understand it) was based on his lack of recollection at
that time about counter injections. Nevertheless, I would add that I can
understand how he failed to remember the counter injection capacity. First, it
related to Legacy Horizon, a system that had not been in operation for six years
Page 466 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
at the point he was working on his statement, and within that, injections via the
counter were (as he appears to have late understood it) a rarely used process.
Second, even for a technically proficient witness, the subject of remote access
was a complex one and the above matters were on a highly technical point within
that already complex subject matter.
856. Based on the documents that I have reviewed for the purposes of preparing this
statement, I can see that WBD received an email from Gareth Jenkins at 12:28
on 16 November 2018, being the day Parker 1 was served, which described this
method of injection: see §§824 ff above. When WBD received Gareth Jenkins’
email, it would have been preferable for us to have asked Stephen Parker about
his knowledge of counter injections. However, Parker 1 was due for filing on that
same day, and WBD was stretched and working under significant pressure of
time. Given the time pressure, the technical subject matter, and the fact that
Gareth Jenkins was providing this information in relation to a different question
being addressed in AVDB2, not Parker 1, I believe that it was therefore
understandable in the circumstances for WBD not to have connected the dots so
as to ensure that the relevant information was raised with Stephen Parker.
857. There were two inaccuracies in Parker 2 regarding remote access that I address
below:
857.1.The first error, which I have alluded to above in the passage quoted at
§847, is that a footnote at paragraph 35 of Parker 2 had stated Stephen
Parker’s understanding that a GIRO transaction was an ‘automated
payment’ (AP) as opposed to a manual bank transaction. My best
understanding (based on my heavily faded recollection of these technical
matters) is that the significance of this was that “in theory, someone could
Page 467 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
have used a transaction injection in Legacy Horizon to carry out a
transaction such as a GIRO bank transfer’, but if so, and a GIRO
transaction was an AP transaction, it would have been automatically
copied to the relevant client (whereas a copy would not be sent if it was a
manual or ‘EPOSS’ transaction). On a draft copy of Parker 2, Stephen
Parker had said “you'll need someone like Gareth to give you a definitive
answer, it was his idea after all. I think the answer is that Giro bank is
ALSO an AP transaction ...”.827 An enquiry was sent to Gareth Jenkins and
on 29 January 2019,8%8 whilst waiting for Gareth Jenkins’ response,
Stephen Parker was asked to review his witness statement and “if he was
happy with it’, to sign it.62° He did so. The following day, 30 January 2019,
Gareth Jenkins responded to WBD’s query clarifying that “the Giro
transactions are not AP” (emphasis added).5°° On 4 February 2019, WBD
emailed Stephen Parker his signed witness statement and informed him
that Gareth Jenkins had explained that the footnote to paragraph 35 was
incorrect and a correction would need to be made.*®*" It is unfortunate that
Gareth Jenkins did not respond in time for the correction to be made to
Stephen Parker’s second witness statement. However, the statement that
a Giro transaction was an AP transaction was one which was made to the
best of Stephen Parker's own knowledge as at that time, and he signed
Parker 2 on the basis that he was happy with it.
827 WBONO0001401.
828 WBONO0001402.
828 POL00363893.
830 WBONO000168.
Page 468 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
857.2.Second, paragraphs 29 and 30 of Parker 2 provided that 14 instances of
counter injections had been identified in response to searching Peak
records in accordance with the search criteria set out in the statement.
The statement described these injections as happening while Richard Roll
was employed at Fujitsu. While preparing for trial, the legal team (I do not
know whether this was originally Counsel or WBD) identified that the
identified Peaks spanned the life of legacy Horizon, rather than just the
period during which Mr Roll was employed by Fujitsu. This was raised
with Fujitsu on 7 March 20198? and on 11 March 2019 Fujitsu confirmed
that this was an error.®°3 Parker 2 was thereafter corrected to state, “during
the time of legacy Horizon” as opposed to during the course of Richard
Roll’s employment with Fujitsu.°4
What more POL and WBD could have done (Q101.4)
858. My principal reflection in relation to the inaccuracies in Torstein Godeseth's and
Stephen Parker's witnesses’ evidence which I explained above is that it would
have been preferable to have started proofing the Fujitsu witnesses earlier about
remote access. This would have given my team greater opportunity to bottom out
some of these highly technical points over a longer period of time, which may
have helped to avoid in accuracies in the evidence. I do not recall giving this
specific thought at the time because I believed that we had bottomed out the
position through Deloitte's work. My other reflection is that it may have been
helpful had there been a further matter Partner instructed at the time to focus
82 WWBON000020200001.
833 WWBONO000210.
834 WBON0001473.
Page 469 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
solely on the HIT. I do not think this was necessary for the duration of the
litigation. However, the rapid litigation timetable at this point, with a significant
amount of evidence for the HIT falling due during the CIT, inevitably put me and
my team under pressures that were extreme even in the context of heavy
commercial litigation.
859. The bigger, persistent, problem, as I have alluded to above, was the Fujitsu
witnesses that did give evidence simply did not have sufficient depth of
understanding of some of the highly technical IT processes that were at the heart
of the trial. Whilst they all undoubtedly were technologically minded people and
were generally well-versed in Horizon, the level of detail into which it was
ultimately necessary to descend to resolve the Horizon issues posed a challenge
even for them. The result was that despite my team’s best endeavours in trying
circumstances, some of the evidence of the Fujitsu witnesses was inconsistent,
inaccurate, and incomplete.
860. As to the broader question of what more WBD could have done “to ensure the
position on remote access was presented accurately from the outset”, I have
reflected carefully while preparing this witness statement, and, in my view, I do
not consider there was much more WBD could have done:
860.1.We (including me) persistently asked Fujitsu searching questions about
the information they provided on remote access. We sent detailed
questions and notes in writing, and sought their employees’ comments.
When their employees provided comments, we did not assume that was
sufficient and consistently asked for further information or more detail.
Page 470 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
860.2.In accordance with POL’s instructions, in 2016 WBD instructed Deloitte in
order to have an expert probe the information which was provided by
Fujitsu on remote access. Throughout this engagement we liaised with
Deloitte and we also asked them detailed questions. However, Deloitte
could also only work from the information that was provided to it by Fujitsu
personnel and from the Fujitsu documents.
860.3.We prepared the LOR and Generic Defence with great care based on what
we then knew. Months of planning and work went into these documents
and they were supported by detailed factual investigations. We went to
significant efforts to ensure that they clearly and transparently POL’s
understanding of remote access, including by seeking Deloitte and
Fujitsu's sign-off.
861. As for what more POL could have done, I consider that this is ultimately a
question for POL.
(vi) Nature of WBD’s involvement in Dr Worden’s evidence (Q96, Q102)
Noting paper on Worden 1 (Q96)
862. In a Noting Paper to the Steering Group dated 28 November 2018
(POL00006471), I provided an update about Dr Worden’s report, Worden 1,
which was due to be served just over a week later. I explained as follows:
“RW's central conclusion is that Horizon is reliable and extremely unlikely
to be the cause of the Claimants’ shortfalls. Much of the debate between
him and the legal team is how best to convey this to the Judge. The legal
team prefer qualitative analysis ("Horizon is sound because...") whereas
RW prefers a quantitative analysis ("The chances of there being a bug
are X%'). RW says that as an engineer he would always base risk
assessment in statistics and judging the "extent of bugs in Horizon" is a
Page 471 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
form of risk assessment. The legal team's concerns with placing too
much weight on statistical analysis are that:
(a) Some judges just do not like numbers and Mr Justice Fraser
may ignore them.
(b) Calculating statistics requires making some assumptions and
those assumptions can always be attacked in cross examination.
3.1.2 This topic has been the subject of numerous calls and conferences
between RW and Counsel (we must have spent over 20 hours debating
this point). The outcome is that:
(a) The statistics will be kept, but only after the qualitative analysis
has already reached a freestanding conclusion that Horizon is
sound. This is why in RW's report section 7 (qualitative) goes
before section 8 (quantitative).
(b) RW has adopted extremely conservative (anti-Post Office)
assumptions so that the assumptions would need to be massively
wrong to move the end-result.
3.2 4 The remote access section of the report needs more work (section
11). RW's view is that it is so obvious that Fujitsu would not abuse their
remote access capability, and that doing so would be so difficult, that the
point warrants not much comment. We have explained the sensitivity of
this topic to Post Office and the case, and he is going to do more work
on this.”
863. In my view, it was proper for WBD to explain to Dr Worden the risks of engaging
solely in statistical analysis. The problem for the legal team, which was a
legitimate one when dealing with statistical analysis, was that the force of Dr
Worden’s conclusions might be lost to a Court if they were presented in purely
quantitative terms. As I explained in the noting paper, some judges are not
persuaded by a purely numerical approach, and we were concerned that this
could be the case with Mr Justice Fraser.
864. The Counsel team and I had a number of discussions with Dr Worden about the
importance of conveying his conclusions in an understandable way. To the best
Page 472 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
of my recollection, this included one or two full days of conferences at Counsel's
Chambers as well as several telephone calls with Dr Worden. As a result of these
discussions, Dr Worden understood that while his reliance on an exclusively
statistical analysis might be persuasive to him because of his background and
the nature of this expertise, it might not persuade others. Dr Worden nevertheless
wanted to maintain a heavily statistical approach, which was his right, but also
included a qualitative assessment in his report.
865. The other concern which the legal team had in relation to Dr Worden’s draft report
was his “worst-case assumption” methodology. According to this methodology,
Dr Worden concluded that at worst, there had probably been 672 bugs in Horizon
over the course of its 18-year history (factoring in the risk of there being latent
and undiscovered bugs). To Dr Worden, this was a tiny number, namely, 1% of
the total number of bugs that he thought that the Claimants would have needed
to prove in order to succeed with their case. The problem was that others (who
did not have Dr Worden’s statistical mind) may have felt this to be a significant
number.. We explained these concerns to Dr Worden in conference and on calls.
But, as I explained in this Noting Paper, Dr Worden “has a duty to the Court to
give his fair opinion and he is giving that priority - as he should — rather than only
saying things that make Post-Office’s position and Counsel's job easier’
(emphasis added).
866. To be clear, the legal team did not draft Dr Worden’s report for him. Counsel and
the WBD team provided comments on the draft report, and we discussed it with
him. As the Noting Paper explains, the substantive conclusions which Dr Worden
had reached were positive, entirely his, and he was very confident about them.
The only concern which the legal team had was how those substantive
Page 473 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
conclusions could be most clearly expressed. In my view, it is proper, and usual,
for the legal team to be concerned with such matters of form and expression.
867. The risks associated with Dr Worden’s approach were communicated to POL
through this Noting Paper. In addition to this, I recall discussing them with Rodric
Williams on several occasions. My views of these risks were as set out in the
noting paper, namely, that 600+ bugs to a layman sounded like a significant
number and could be misreported by the media. In the event, these risks were to
be managed by POL’s Counsel Team (who had the task of explaining Dr
Worden’s assumptions and conclusions in an understandable and persuasive
way), and by POL’s communications team — in relation to which I had no
involvement.
Involvement in Worden 3 (Q102,
868. The Horizon Issues Trial was paused for a period from 21 March 2019 due to the
Recusal Application (which I deal with below in Section Q). During this
adjournment, Dr Worden came up with a new idea for how to analyse the risk of
bugs affecting branch accounts. On 27 March 2019, he wrote:
‘I'd like to give you an early heads up on a type of analysis I have been
doing recently, which I think is very promising.
It can give me another independent way to estimate the 0.4% upper limit
on the impact of bugs on claimants, that can be done quite quickly as
follows:
If there is a bug affecting a branch's accounts, the branch FAD code is
quite likely to appear in a Peak (as is confirmed by Callendar Sq,
Receipts Payments mismatch, suspense account). Say the probability is
as low as 50%; it is probably higher, for any significant financial impact.
Out of 218,000 Peaks, there are about 1,700 which mention the FAD
code of one or more claimants, at a date when the claimant was in post.
(I have written a program to find them) To calculate the total impact of
Page 474 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
bugs on all claimants accounts, we just have to sum the likely financial
impact for each Peak that indicates a bug, over those 1,700 Peaks.
To get a total impact of £18.7M, the average impact on a branch per
Peak has to be £10,000 (or £5000, if you allow for the 50% factor above).
This is very high indeed, for an impact on one branch. PO would have
been going mad about it.
For most of the 1700 Peaks, it is pretty obvious that they had no impact
at all, and were not bugs they are about ISDN or something irrelevant. It
would be quite simple to examine a random sample of 100 Peaks out of
the 1700, and scale up by 17.
I have not done this, but I will bet that the resulting number is tiny -
probably less than 1% of £18.7M
This would give three very independent upper limits for the proportion of
claimants' losses arising from bugs - which will probably be 8%, 0.4%
and say 1%.
These limits come respectively from the claims data, KELs, and Peaks.
Three very different sources and assumptions. It will be very hard for the
Cs to get away from these three independent analyses - i.e. to prove that
they are all wrong.’®°>
869. On the same day, I forwarded Dr Worden’s email to Tony Robinson QC, who
responded as follows:
“My immediate reactions to this suggestion are as follows:
1. on its own, the sheer number of peaks referring to claimant branches
(1700) looks very bad for us;
2. so far, there has been no suggestion of any evidence of a bug affecting
any claimant branch, this exercise appears to give rise to such a
suggestion, why would we ever want to do that;
3. it feels dangerous even to mention these peaks unless and until
we have read them all and satisfied ourselves that they do not
record any bugs affecting claimant branch accounts;
835 WBONO0000701.
Page 475 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
4. putting the last point the other way round, if we had read them
all and satisfied ourselves of this, I would love to talk about them,
but not before;
5. RW suggests a sampling exercise, but what if a significant proportion
of his sample looked as if they might be claimant-branch-affecting bugs;
6. we already have too much maths, too much averaging and too
many assumptions about what it would take to make Post Office go
mad;
7. I doubt that this new calculation will help much (in the sense of
establishing something that might cause the judge to take a more
positive view of our case) and I strongly suspect that it will do more
harm than good (in the sense that the optical problems which it will
create for us as discussed above will be greater than any help it
might give us); and
8. this is the sort of exercise to be done in breach trials, not this
trial;
9. it troubles me that RW has already found 1700 peaks without
discussing it with us first — thereby creating a risk that if Green gets ina
lucky question, RW could be required to talk about them in cross
examination; and
10. what the hell does RW think he is doing moving the goalposts
in this dangerous way this close to the endgame’®* (emphasis
added).
870. Junior Counsel, Owain Draper, responded:
“If the results are good, RW will be slaughtered for raising it so late, and
the Judge would attach very little weight to it. It’s almost a one-way bet
against PO in that bad results would be awful.
It’s already a big problem that RW has identified the Peaks. If he now
does no work on them, it will look awful if it comes out. If he does the
work and we do not provide it to Cs, it will look awful if it comes out.
/'ll think further on it, though, because it seems counterintuitive that all
available options are so bad!”63”
896 WWBONO000702.
837 WBON0001536.
Page 476 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
871. I shared these concerns. As I explained in an email to Rodric Williams dated 6
April 2019, there were three major concerns with Dr Worden’s proposed new
methodology:
“It will look like a last minute ambush. [Mr Justice] Fraser has complained
bitterly about this type of conduct by experts in previous judgments
outside this litigation.
If the results are good for PO, it will look like the legal team has pushed
this.
[Patrick Green QC] will ask questions about why this analysis has not
been done before (to which RW will say that it only recently occurred to
him to do it).
We will necessarily need to give Coyne the right of reply.
In our view, even if the content of Worden 3 is good for PO (and that is
not guaranteed), it will backfire and lead to criticism from the Court.
The above has been discussed at length between RW and Counsel. We
have pressed RW hard not to do this work or produce Worden 3. RW is
resolute that he is required under his expert duty to bring this new
information to the attention of the Court. There is no prospect of
persuading him not to do it and we obviously have no power to block
this”,658
872. There were three further problems. First, there was no time to conduct the in-
depth analysis required of the Peaks that Dr Worden referred to in order to be
satisfied as to what evidence they actually contained. Second, it was more
statistics, and the legal team were already concerned about the amount of
reliance Dr Worden was placing on this kind of analysis. Third, it seemed to add
little to Dr Worden’s original analysis. On balance, the legal team’s combined
view was that this new analysis was likely to do more harm than good. I recall
838 POL00042611
Page 477 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
that HSF also shared these concerns, when in due course they were instructed
by POL in mid-April 2019.
873. However, Dr Worden genuinely felt that his new report was valuable; he had done
additional work, and that additional work therefore fell to be part of his overall
opinion. We could not tell him not to do it and, once done, the analysis would
need to be shared with the Claimants and, consequently, the Court.
874. By 28 March 2019, Dr Worden had already drafted a letter to Freeths to inform
them of his new analysis. I explained that my instinct was for him to complete the
analysis before engaging with the Claimants or Mr Coyne.®*9 In my view, it would
have been inappropriate for Dr Worden to put forward his proposed additional
evidence to the Claimants and the Court without even completing his analysis.
However, I was always mindful of Dr Worden’s duty to act independently and his
duties to the Court and I respected the fact that he had done this additional work.
Dr Worden remained adamant that he should finish and submit his new analysis.
On 6 April 2019, he wrote to me in the following terms:
“As you know, I believe it is my expert duty to do this, so I am going to
do it. But I am acutely aware that having done my ‘independent expert’
thing on you, you have to manage the fallout in all directions. You are
going to have to explain it to PO on Monday, stage-manage the whole
presentation, etc. I'd like to say why (in my view) even if the short-term
fallout is a pain, in the longer term it will be a very good thing for PO.
Maybe not in the best order:
The trial is all about bugs in Horizon, and how they might have impacted
the 560 claimants
839 POL00112051
Page 478 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Over 20 years, Fujitsu have kept very good records - in the Peaks,
OCRs, MSCs etc - of any possible bugs in Horizon, discrepancies,
remote access events, etc.; and what branches they affected
Fujitsu could never be accused of any ‘corporate cover-up' of defects in
Horizon. They wanted Horizon problem-free, and have done a good job
of keeping it that way. They have chased down any possible bug as soon
as they could, and recorded the process.
So to find out how bugs in Horizon impacted claimants, you just need to
go to those records, filter them by claimants' FAD codes, and count the
possible impacts. This means 2,400 documents, rather than 500,000 .
It's that simple, and it is a small job (why didn't I think of it before??)
Compared to all that has gone before, this is a massive simplification of
the case. Any judge would grab it with both hands , just to simplify his
own job. There is no excuse for taking a complicated and obscure route
- or rather, ignoring the simple route - when such a simple route exists.
That is why I have to tell the court about it.
It is also very bad for the claimants. Coyne and Green thrive on
confusion; it is their only weapon. This removes confusion - or goes right
round it - and cuts off the life-blood of their case”.®4°
875. On 7 April 2019, I informed Tony Robinson QC that Dr Worden’s decision had
“gone down like a lead balloon at PO”.®41 Tony Robinson QC responded:
“I’m not surprised. I imagine that, if we serve a Worden 3, the claimants
will claim that we are acting tactically — they may even link it to our
recusal application (but for that application, there would not have been
time for RW to wrestle with these questions etc).
There is no problem-free way of addressing this issue.
[ol
It seems to me that his ever-expanding views as to the number of
Horizon bugs which were branch-affecting do count as changes of view,
but that could probably be addressed in the course of the expert
meeting that the experts have agreed to have anyway.”*2
840 POL00042614
841 WBONO0000703.
842 WBONO0000704.
Page 479 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
876. On 8 April 2019, Owain Draper emailed to advise that Dr Worden may not get
permission to adduce another report but “/ do not think we can stop RW doing
what he wants to do, even though it may well mean the court refusing permission
and getting angry. We have already told him that we do not think the further
analysis changes much or is even within a Horizon Issue”.*3 Owain Draper
advised that Mr Coyne be informed as soon as possible about Dr Worden’s
position.
877. On 10 April 2019, WBD wrote to Freeths to inform them about the potential new
report. The letter stated “Please note that this came as some surprise to us and
we wish to make it clear that neither our client nor its legal team requested this
further work, the necessity for which we have discussed with Dr Worden, but
which Dr Worden decided to undertake pursuant to his understanding of his duty
to the Court.”*4
878. Tony Robinson QC was due in court on 10 April 2019 and wanted to know his
instructions on whether he should be seeking permission for supplemental
reports. POL’s instructions were leave it to the court to decide what to do next.®45
Mr Justice Fraser’s decision was to make an Order directing that the experts
should meet one more time.
879. Pursuant to Mr Justice Fraser’s Order, Dr Worden sought to engage with Mr
Coyne to discuss the new analysis but Mr Coyne was resistant, in part due to
other commitments.*°
®43 WBON0001539.
844 WWBONO0001543.
®45 WBONO001547.
®46 POL00112145.
Page 480 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
880. By 25 April 2019, Worden 3 was largely finalised and sent to Mr Coyne in draft
on a without prejudice basis.°4”
881. On 26 April 2019, Dr Worden sent me an email setting out a number of reasons
as to why he thought POL should seek to rely on the report. Dr Worden
considered that, even if permission was refused, all possible outcomes were
good for POL:
“Now we come to the main reason to make the application - which I
believe PO should consider before you and they decide.
Judge has had Coyne's and my reports for several months. He can
surely see the difference between Coyne's anecdote-based approach
and my numbers-based approach. We do not yet know which approach
he prefers, or why.
Sending report 3 to him, with an application, will be a litmus test of his
attitude to numbers. If - as you all suppose - he hates numbers, he will
reject the 3rd report. If, as I suppose, he is a bit of a geek and fancies
his techie expertise, he will not dismiss it out of hand - and may welcome
a simpler route to deciding the issues.’*8
882. I shared this email with Tony Robinson QC*® who disagreed and advised
that Mr Justice Fraser would characterise the application as an exercise in
oppression by POL:
“In this context, our concern is not that the judge will find that Robert is
a PO stooge (although that is what the judge may be hoping to find it his
final judgment and it is always possible that he may rely on this exercise
as one of his grounds for doing so). Our concern is that the judge will
characterise our application to rely on the new report as an exercise in
oppression by us.”®°°
®47 WBONO0000708
848 WWBONO000710
®48 WWBONO000711
850 WWBONO0000712
Page 481 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
883. I began to prepare an advice to POL about what they should do about Worden
3. By this point the view was that the Peaks analysis in the report was broadly
unhelpful to POL but the remote access analysis was helpful. Counsels’ view was
that seeking permission to rely on the remote access section only was the best
option, followed by seeking permission to rely on it all (despite the unhelpful
parts).
884. On 30 April 2019, I advised Rodric Williams of Counsels’ view.®>' We suggested
writing to Freeths in the first instance seeking their views so as to be in a better
position to gauge the risks. I can see from my calendar that there was then a call
to discuss this issue with WBD, Counsel, Rodric Williams and HSF on 1 May
2019. My recollection was that Rodric Williams and HSF also considered this a
very difficult problem to solve.
885. On 3 May 2019, WBD wrote to Freeths to inform them that Dr Worden felt it was
his duty to update his report and POL were considering their position on whether
to apply for permission to rely on a supplemental report.®? Freeths replied
complaining about Dr Worden and POL’s approach.®°° WBD then gave further
advice to POL (and HSF who by that point were tasked with seeking instructions
from the Board Subcomittee as and when they saw appropriate) on how to
proceed on 16 May 2019.5°4 We were concerned that Freeths were attacking
Worden 3 on the basis that it undertook Claimant-specific analysis, but similar
analysis was in Worden 1 and 2; and if Worden 3 was rejected as being out of
scope for the Horizon Issues trial then we could face similar problems with
851 WBONO001578.
852 POL00274897
853 POL00274899
854 WBON0001585.
Page 482 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Worden 1 and 2. Therefore, even though the remote access section was
valuable, Counsel advised that POL should not apply for permission to rely on it.
The end advice, therefore, was that Dr Worden should send the third report to
Jason Coyne on an open basis to see if Jason Coyne engaged. If not, Worden
3 would be sent to the court the next Wednesday. HSF commented that this was
a sensible plan.8°5
886. I advised POL on 22 May 2019 that Dr Worden would write to the court with the
report that day.®5* We maintained our advice of 16 May 2019 that POL should
not apply for permission to rely on the report, and HSF agreed.®°’ We obtained
instructions from POL not to apply for permission to rely on the report.®©°
887. Dr Worden emailed the court with his report®° and Katie Simmonds (of WBD)
provided him with recommendations on the drafting of his email.®®° This then
resulted in Mr Justice Fraser replying to say that witnesses were not entitled to
communicate directly with the Court, as reflected in Dr Worden’s email to me (to
which Q102 of the Request refers, i.e. POL00112279).
888. Worden 3 was discussed in Court on 23 May 2019 and POL was directed to file
a witness statement explaining the chronology of its development, which I did on
31 May 2019. This was my seventeenth witness statement.®®"
889. On 28 May 2019, WBD advised Freeths that POL did not intend to use Worden
3 in cross-examining Mr Coyne, or to seek permission to rely on it, but that (i)
855 WWBONO0001590.
856 WBONO0001600.
857 WBONO0001600.
858 POL00042688.
858 WWBONO0007 14.
860 WWBONO001607.
861 POL00275716
Page 483 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
ways of analysing the evidence based on samples may be discussed with Mr
Coyne and (ii) Dr Worden’s answers in cross-examination may need to refer to
the contents of the report if he was to comply with his obligation to tell the truth.®©2
By correspondence dated 31 May 2019, the matter was drawn to a close.®©>
890. On the question of whether I ‘advised’ Dr Worden to contact the Court directly:
as the contemporaneous documents clearly demonstrate, this was an
unprecedented situation for me (and I believe for Counsel and HSF also) and it
was difficult to know how best to respond to it. My view, and what I understood
to be the view of Counsel, was that no matter what approach we took to the
situation, it was likely to reflect badly on POL (through no fault of POL’s own).
891.1 do not recall whether I advised Dr Worden to contact the Court directly.
However, upon my review of the contemporaneous documents for the purposes
of preparing this statement, the evidence indicates that Dr Worden came to the
decision that he would contact the Court directly of his own volition, and after he
had made that decision and informed us of this, we then assisted him to draft his
email after taking instructions from POL.
892. As to this, CPR 35.14(1) permits experts to “file written requests for directions for
the purpose of assisting them in carrying out their functions”. Dr Worden was
aware of this provision (the Managing Judge had expressly raised it with the
parties the previous year, had made reference to it in a CMC Order,®* and Mr
Coyne had written directly to the Court to seek directions in relation to requests
for further information in relation to the Horizon system); he was aware of his
862 WWBONO0001633.
863 WBON0001640.
864 WBONO0001232 at paragraph 13.
Page 484 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
duty to the Court; and he was adamant that in order to comply with that duty, he
needed to inform the Court about Worden 3.
893. On 30 April 2019, I received an email from Jonny Gribben which stated:
“When I spoke to Robert late yesterday afternoon he mentioned that
he will probably ask the Court for directions under CPR 35’°°
(emphasis added).
894. Tony Robinson QC responded to this, stating “That’s an important part of the
jigsaw which we need to know’.®® I then responded:
“And it’s good that RW has reached that view on his own. For the
record, I’ve not mentioned this possibility to him or anything about
our plans for Worden 3. Should we ask what directions he might seek? I
can do that with an open question so that we are not leading him’®®’.
895. WBD and the Counsel Team considered that Dr Worden was permitted under
CPR 35.14 to take this course of action. On 23 May 2019, both parties appeared
before the Court in order to address Dr Worden’s email. The relevant exchange
between Leading Counsel and Mr Justice Fraser states:
“25 My Lord, to deal with your Lordship’s second point,
1 in my respectful submission it is entirely proper for
2 an independent expert to communicate directly with the
3 judge. That underpins CPR 35.14, which your Lordship
4 will also be well aware of, the provision which allows
5 experts to seek the directions of the court.
865 WWBONO0000713
866 WWBONO0000713.
887 WBONO0000713.
Page 485 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
MR JUSTICE FRASER: Well, whether that’s correct or not, in
7 the circumstances of this case, the way in which it was
8 done and the time at which it was done -- and I went
9 back and reread what you had said to me on 11 April - -
10 two points arise. I’m_not in any way finding or stating
11 that it was improper for him to have done that.
12 MR DE GARR ROBINSON: I'm grateful for your Lordship to say
13 that.
14 MR JUSTICE FRASER: However, it is highly unusual for
15 an expert to do that without some sort of prior
16 notification that that’s happening; and secondly, it was
17 not clear on the face of his email whether the claimants
18 knew that was happening.
19 MR DE GARR ROBINSON: My Lord, there was prior notification.
20 MR JUSTICE FRASER: Right” (emphasis added).°®@
896. Although Mr Justice Fraser was critical in his Horizon Issues judgment (at [725]-
[726]) about Dr Worden having directly emailed the Court, as the above transcript
illustrates, Mr Justice Fraser’s view at the material time (the day after the email
was sent) was that it was not “in any way ... improper’ for Dr Worden to have
emailed the Court per se, but rather, for him to have done so “without some sort
of prior notification” and without the Claimants knowing what was happening.
Unfortunately Dr Worden had omitted to copy the Claimants’ solicitors on his
email to the Court, but did separately forward it to them three minutes later, which
868 POL00042714.
Page 486 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
was the source of the Court's understandable concern.®? Leading Counsel
clarified the position with the Court and explained that there had been proper
notification to the Claimants.
897. Leading Counsel also invited the Court in closing to provide guidance on how to
approach a similar situation in the future. Mr Justice Fraser concluded in his
judgment that he did not know why “it should be thought necessary to tell
witnesses — lay or expert — that they should not unilaterally communicate with
the court’ (at [728]). With respect to Mr Justice Fraser, Leading Counsel invited
the Court to provide this guidance because this was a difficult, unprecedented
situation, and the legal team did consider it to be clear as to how POL ought best
to have dealt with it.
898. In conclusion, in relation to Q102.2, I do not believe that I advised Dr Worden to
write to the Court before he independently came to the view that this is what he
would do. When he reached that decision and informed us of this, we assisted
him to do so, because we (WBD and Counsel) considered that he was entitled
to do so under CPR 35.14, and, most significantly, Dr Worden considered that
he was. This explanation also addresses Q102.3.
(vii) The Board Subcommittee meeting 21 February 2019 and advice on merits
(Q100)
899. Q100 refers me to the minutes of a meeting of the Board Subcommittee dated
21 February 2019 (POL00006753) record a “brief[ing]’ by Post Office’s Leading
Counsel, Tony Robinson QC. Prior to this meeting, I prepared the first draft of a
869 WBONO00007 14.
Page 487 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
speaking note for Tony Robinson QC which he then reviewed and amended.®”°
Additionally, the day before the meeting, WBD circulated a risk assessment table
which we had shared with Tony Robinson QC in advance.8”' Jane Macleod (then
General Counsel) sent this risk assessment table to members of the POL Board,
namely Tim Parker, Ken McCall, Tom Cooper, Paula Vennells and Alisdair
Cameron.*”2 The minutes need to be understood in the context of these two
documents.
900. I do not recall who produced the minutes of the meeting and searches of my
email records indicate that I did not receive them at the time (cf. Q100.2). I
strongly suspect that the minute taker was Veronica Branton (POL), as she had
been a minute taker at other meetings.
901. My recollection is that Tony Robinson QC spoke for around 30 minutes and
closely followed the speaking note, before inviting questions. The points to be
taken from the speaking note are as follows:
901.1.The purpose of Tony Robinson QC's briefing was to explain to the Board
Subcommittee the case which POL was going to advance at trial.
901.2.Tony Robinson QC’s assessment was that “the evidence weighs in favour
of Horizon being a robust system. There is little risk of the Judge saying
that Horizon is bad but he may find that it is only “ok” [and]... This trial will
come down to whose expert is more credible. We believe the evidence of
our expert, Dr Worden, has a better methodology and is more cogently
evidenced that the Cs expert, Coyne.” Tony Robinson QC also noted that
870 WBON0000337; WBON0001421; POL00112903.
871 WBON0001418.
872 POL00024150; POL00265865.
Page 488 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
there was some judicial risk that needed to be borne in mind “as this Judge
as [sic] acted in ways that are truly extraordinary as you will have already
heard from David Cavender on the Common Issues trial’.
901.3.Tony Robinson QC summarised Post Office's core argument as follows:
“Horizon like all IT systems is not perfect but it is a very good system...This
is not about proving that Horizon is perfect and that there are no bugs in
Horizon...In the context of this litigation, it is about showing that Horizon
accurately records transactions the vast majority of the time so that PO,
SPMs and the Court can safely start from an assumption that the branch
accounting information held on Horizon is sound...No-one has found a
fundamental flaw in the system or its support processes. We are nowhere
near a situation where this a bad or even average system. This is a
decision between whether Horizon is good or very good...”
901.4.Tony Robinson QC summarised Dr Worden’s “clear and well-explained
opinion” as follows: “the volume of bugs Cs have found is tiny. About 20
bugs over an 18-year period, against a back drop of 30,000 active users
and 50m transactions a week. Even if one assumes that the Cs case is
entirely correct, Horizon would still be a robust system.”
901.5.Tony Robinson QC explained that “the strength and formulation” of Post
Office's case “may need to be adjusted in light of the Cs supplemental
expert report served on 1 Feb this year...On 1 February, [the Claimants]
served their supplemental report. It is anything but supplemental. It raises
a whole new range of issues, including about 15 new alleged bugs in
Horizon that have not been raised before. The legal team and FJ are now
working ferociously to investigate these new points. Progress so far has
Page 489 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
been good and so we are hopeful, but not certain, that we will have
credible counter-points to raise to most of Coyne's new attacks”.
901.6.Tony Robinson QC then summarised the key risks:
(i) POL’s case that Horizon is robust and extremely unlikely to cause
shortfalls has set a “high bar’. This bar was set based on Deloitte’s
findings and the fact that POL’s business operations were built on
this basis. Although this is a high bar, it is one which “Dr Worden ...
felt comfortable” giving an opinion on, and it aligned with Deloitte’s
findings.
(ii) Dr Worden’s evidence relied heavily on statistical models which
some Judges may not be receptive to, but this risk had been
mitigated by Dr Worden’s inclusion of a qualitative opinion alongside
his quantitative analysis.
(iii) The main factual evidence for Post Office would be given by
technical personnel from Fujitsu. In a small number of cases, the
evidence given was incorrect and needed to be corrected through
further statements. This has given rise to the concern that further
cracks may appear under cross-examination. This is not due to a
lack of cooperation by Fujitsu, but rather that they were being asked
to give factual evidence on issues that arose sometimes decades
ago on highly technical and often obscure points. These risks had
been mitigated through extensive discussions with Fujitsu, probing
Fujitsu’s evidence as much as it is possible for a lawyer to do, and
Tony Robinson QC had met with Fujitsu on a number of occasions.
Page 490 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
(iv) “Weaved into the heart of the Cs case [is] the conspiracy theory that
PO and FJ are meddling with branch transaction data in secret and
that this is causing the shortfalls and branches”. The difficulty Tony
Robinson QC identified was that POL’s and Fujitsu's case on remote
access has changed over the years. While he thought it added little
substance to the Claimants’ case, it could prove damaging. Although
there was no evidence of widespread data manipulation, and
although it was difficult to imagine any rational reason why Fujitsu
would maliciously or carelessly manipulate branch data, Fujitsu’s
changing position on remote access exposed Post Office to
allegations of a cover up. This could then be used to tarnish the
credibility of Fujitsu’s witnesses.
902. As for whether there are any material inaccuracies in the minutes (cf. Q100.4.2),
the phrasing of the minutes in paragraph 2, section 2 makes it sound as if Tony
Robinson QC is advising on the merits. In fact, as can be seen from the speaking
note, Tony Robinson QC was setting out POL’s case theory for the HIT for the
benefit of the Subcommittee.
903. What the minutes do not accurately reflect is the qualification which Tony
Robinson QC had given, which was that the strength and formulation of POL’s
case would depend upon the Claimants’ supplemental report (Coyne 2) which
raised a number of new issues, including around 15 new alleged bugs, and POL’s
ability to respond to it. This point was also emphasised in the risk assessment
table prepared by WBD, the relevant passage of which stated:
“In our view, the report of Dr Worden adopts a better methodology and
is more cogently evidenced but it remains open challenge in Court. This
Page 491 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
view is subject to our introductory comments about Post Office providing
adequate responses to Mr Coyne’s new arguments in his Supplemental
Report. Mr Coyne’s new points have been reviewed by Dr Worden who
is unmoved in his opinion”.8”3
904. I do not recall Tony Robinson QC using the words “critically robust’ (cf. Q100.3)
and I note that these words are not contained in his speaking note.®’* However:
904.1.The speaking note states that “the evidence weighs in favour of Horizon
being a robust system’.
904.2.The speaking note summarises Dr Worden’s evidence as supporting the
view that, even on the Claimants’ case, Horizon is a robust system.
904.3.The risk table explains that the evidence of Dr Worden was that “Horizon
has been a very robust system, compared to other major systems I have
worked on in sectors such as banking, retail, telecoms, government and
healthcare” and that “the robustness of Horizon made it extremely unlikely
to be the cause of shortfalls in branches”.
904.4.The risk table concludes that “the report of Dr Worden adopts a better
methodology and is more cogently evidenced but it remains open [sic]
challenge in Court.”
905. In relation to whether the sentence “For the vast majority of the time, Horizon
was a very reliable system” was materially accurate, the speaking note states,
“Our case ... The core argument we will run in Court is ... that Horizon accurately
records transactions the vast majority of the time so that PO, SPMs and the Court
can Safely start from an assumption that the branch accounting information held
873 POLO0265865.
874 WBON0001422.
Page 492 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
on Horizon is sound.” Again, this sentence was not in the nature of advice about
the merits of POL’s case, but an explanation of the case which POL would
advance at trial.
906. In relation to the sentence, “the bar we have set ourselves was very high as we
had said that the Horizon System was robust and very unlikely to cause
significant losses. We had to be able to support this starting position. Not meeting
that bar would have a serious impact on PO Limited's operating procedures and
would open up 18 years of previous decisions’, the relevant section of the
speaking note states (and it is worth setting it out in full):
e “PO has necessarily set its case very high: that Horizon is robust and
extremely unlikely to cause shortfalls. It has done this for two reasons:
e First, this was the finding from Deloitte's investigation into Horizon at the
outset of the litigation and on which PO premised its legal position.
e Second, PO's operating models are built on the assumption that Horizon
works.
o The heart of this litigation is whether SPMs should be liable for
shortfalls. The Cs main line of attack is that the shortfalls were not
caused by them but by bugs in Horizon.
o In its interactions with SPMs, PO's starting assumption is that
Horizon works. It habitually looks first for errors in branch before
looking at Horizon. Its contracts with SPMs are structured to [sic] in
line with this approach.
o This approach can only be sustained if Horizon is so reliable to justify
a starting, but rebuttable, assumption that Horizon is accurate.
o. If Horizon does not meet this bar, then that may require PO to test
the accuracy of Horizon in relation to every branch shortfall going
forward before seeking to recover any monies.
co It would also pose a risk to its decisions over the last 18 years in
relation to the recovery of shortfalls and the termination of SPMs:
which heightens the possibility of successful claims within the
litigation.
Page 493 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
e Although this bar is high, Dr Worden has felt comfortable giving an
opinion saying that Horizon is robust and extremely unlikely to cause
shortfalls in branches, which aligns with Deloitte's earlier work.”
907. With reference to Q100.4.1, my views on the prospects of success in showing
that Horizon was robust and unlikely to be the cause of unexplained losses,
mirrored those of Tony Robinson QC. I also agreed with Tony Robinson QC that
the “strength” of POL’s case “may need to be adjusted in light of the Cs
supplemental expert report ... and so we are hopeful, but not certain, that we will
have credible counter-points to raise to most of Coyne's new attacks.” However,
as I set out at §§795-800 above, Dr Worden remained firmly of the view that even
15 more bugs was a very small number, as compared with the number of bugs
which he identified the Claimants would have to prove in order to demonstrate
that issues with Horizon could be a cause of the SPMs’ shortfalls. For these
reasons, I was still fairly confident in POL’s case as to the robustness of Horizon,
but, as the speaking note and risk table demonstrate, I certainly did not think (or
advise) that it was without risk.
908. As to the question whether Counsel or I were concerned about the safety of past
convictions which relied on Horizon data (Q100.4.2), this did not form part of our
analysis. As the minutes, speaking note and risk table make clear, we were
focused exclusively on assessing risk in the group litigation.
909. I do not recall anyone at this meeting expressing any concerns to me about the
safety of past convictions using Horizon data (cf. Q100.4.3).
910. With reference to Q100.5, the Subcommittee received both the risk table and
heard Tony Robinson QC’s briefing, which included a briefing on concerns
relating to remote access. The summary of Tony Robinson QC’s explanation of
Page 494 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
the risks relating to remote access is outlined at §§901.6(iv) above. The risk table
contained the following description of the remote access issue:
“Post Office (via Fujitsu) has always had “remote access” capabilities
and this has been admitted in earlier Court documents.
There is a material dispute as to the extent that “remote access” was
used to alter branch data and whether such access was properly
controlled. Post Office's case is that “remote access” is a rare event and
only used following strict protocols. The Claimants look to paint a picture
of frequent unregulated use of “remote access” to change branch
information in a clandestine manner.
Both experts agree that Post Office had no ability to remotely delete or
edit data within Horizon. The tools for deleting and editing branch data
were held exclusively by Fujitsu. Post Office is therefore reliant on Fujitsu
for evidence of its use and control of these tools.
Fujitsu’s evidence on this subject has been less than satisfactory. During
the mediation scheme, Fujitsu told Post Office that it could not edit
branch data (only that it could inject new transactions). Further
investigations revealed this not to be the case — Fujitsu do have the
ability to edit transaction data by accessing and amending the underlying
database tables within Horizon.
This shifting position has continued during the litigation and has resulted
in Post Office having to file a second witness statement to correct some
errors in the primary evidence of one Fujitsu witness. In light of this, there
are material concerns about whether the Fujitsu witnesses will come up
to proof under cross-examination.
Further, Fujitsu’s record keeping around use of its remote access tools
is incomplete. Some of this is a product of time and document retention
policies, particularly in relation to the old version of Horizon (pre-2010),
but some is due to a lack of structured documentation around the use of
these tools (including a lack of automatic access logging software within
Horizon).
Mr Coyne has placed considerable emphasis on “remote access” in his
Supplemental Report and Fujitsu’s answers to these points will be vital
to the outcome of this issue. Investigations in this regard are continuing.
This is an area where further evidence from Fujitsu, if reliable and
allowed by the Court, would be useful and could affect the overall merits
on this topic.
Page 495 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Even if remote access is possible, it is very unlikely that Fujitsu are acting
maliciously or carelessly causing shortfalls in branches through remote
access. They would have no motivation to do so and the reputational
damage to it of doing so through poor practices would be severe. The
challenge at trial will be persuading the Judge of this and avoiding him
getting drawn into the Claimants’ conspiracy theories.
In terms of impact, the Claimants are seeking a finding that there is
frequent uncontrolled remote access and that that undermines Horizon
being reliable. They are looking for a crossover effect to Category A
above. This link is not however obvious and the Claimants’ expert has
not explained how one issue affects the other. Given the scale of
Horizon, we will argue that the level of uncontrolled access (if any) would
need to be significant for one to lose confidence in the system.
The above factors could lead to a number of different outcomes. For
example, Post Office (or Fujitsu) could be found to have poor access
controls but it be accepted that Fujitsu did not misuse those controls. Or,
that the remote access tools were misused but on such small scale that
there is no overall impact on Horizon or the litigation.
In our view, it is likely that the Judge will make some form of adverse
finding against Post Office on this topic, but it is much more difficult to
assess the impact of that finding on the reliability of Horizon or the
litigation.
Regardless of the outcome, we anticipate that this issue will attract
media attention and poses the greatest risk of reputational harm’.875
911. This section accurately reflected my understanding of the remote access issue
at the time as well as my views as to the risks associated with it.
912. With reference to Q100.5, I do not know if the Board Subcommittee was provided
with the Deloitte reports (assuming this means the Project Zebra and/or the
Project Bramble reports). At the time those reports were produced, the
Subcommittee did not exist (it was created in or around March 2018). Before that
point, the external legal team liaised only with the Steering Group, who then
reported up to the Board and POL senior management. I do not know what other
875 POLO0265865
Page 496 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
documents, beyond the risk table, were shared with the Board Subcommittee. I
did not have a direct line of communication with the Board Subcommittee and I
was not aware of all of its interactions.
Q. RESPONSE TO THE COMMON ISSUES JUDGMENT AND RECUSAL
APPLICATION (Q103 to Q118)
913. In this section I address the Inquiry’s questions at Q103 to Q118, about the steps
POL took in response to the Common Issues Judgment. The Common Issues
Judgment was circulated in draft on 8 March 2019 (the Friday before the Horizon
Issues Trial commenced on the Monday), and was formally handed down on 15
March 2019. In particular, this section addresses my involvement in the
application which POL issued on 21 March 2019 seeking Mr Justice Fraser’s
recusal as the Managing Judge in the GLO (the “Recusal Application”) and,
where relevant, POL’s wider appeal against the reading of the SPMC and NTC
at which Mr Justice Fraser had arrived in the CITJ. As will appear from the below,
I retained overall responsibility for the conduct of the Recusal Application as the
matter Partner for the GLO, although in practice Tom Beezer (another Partner at
WBD) supervised the day-to-day management of parts of this workstream whilst
I was fully immersed in the HIT.
(i) Initial response to the Common Issues Judgment — Summary Note and David
Cavender QC’s advice on appeal and recusal (Q103, Q107 to Q108)
Summary Note (Q103)
Page 497 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
914. By Q103, I am asked to consider POL00022940, which is a six-page ‘Initial
summary of the Common Issues Judgment’ dated 9 March 2019 (the “Summary
Note”). I wrote a draft of the Summary Note at POL’s direction, shortly after I
received the draft Common Issues Judgment the previous morning (as I
understood that POL wanted to begin mitigation planning on the Monday
morning).876
915. Specifically, I am asked to consider a number of excerpts from the Summary
Note.
916. First, I am asked to summarily identify the “large swathes of inadmissible
material’ (paragraph 1.1) that was considered by the Judge in the Common
Issues Judgment. The inadmissible material to which the Summary Note refers
is the evidence led by the Claimants about matters which, in POL’s submission,
went to contractual performance and breach in the six Lead Claimants’ cases,
and in my view was not relevant to the 23 Common Issues which Fraser J had
had to determine. This included matters such as: the adequacy of the training
received by the six Lead Claimants after their appointment; the quality of the
NSBC helpline; POL’s alleged knowledge of problems with Horizon; the quality
of POL’s investigations into the causes of shortfalls; the circumstances of the
Lead Claimants’ suspensions and terminations; and whether POL had sent them
unjustified demands for payment and/or threats of legal action. I have set out
POL’s position on these issues in detail above at §§697 ff.
917. Second, I am asked to identify the “clear evidence of [the Claimants’] dishonesty”
(paragraph 1.4). I do not recall what this sentence referred to, but I assume that
876 See WBON0001446; POL00267481.
Page 498 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
it was a reference to evidence of false accounting by some of the Lead
Claimants.
918. A significant example of this that comes to mind is the evidence Elizabeth
Stockdale gave in the CIT. POL terminated Mrs Stockdale’s contract because
she had submitted a false account, which had the effect of hiding a discrepancy
in her branch. She was asked in cross-examination about whether she had
intentionally misstated her accounts so as to hide discrepancies. Having been
warned by the Judge about her right not to incriminate herself, Mrs Stockdale
declined to answer the question. Mr Justice Fraser decided not to make a finding
on this point; to quote the Common Issues Judgment at [328]:
“I found Mrs Stockdale to be a careful and accurate witness, and I
consider she was telling me the truth. The single question that she
declined to answer was that she had been misstating the accounts to
hide discrepancies. Whether she was right to act as she did at the time
regarding her accounts is a matter for another trial. As with the other
Lead Claimants, I am making no findings in respect of breach, causation
or loss.”
919. Conversely, the paragraphs preceding this point made extensive factual findings
about the shortfalls Mrs Stockdale had suffered and what steps she and POL had
taken in relation to them (see from [302] onwards), including with respect to the
quality of the helpline response, the adequacy of POL’s investigation into the
shortfalls, what options Mrs Stockdale felt were open to her, the conduct and tone
of POL’s correspondence with her, and matters relating to her suspension and
termination in 2016. POL’s case was that Mrs Stockdale’s evidence on these
matters was irrelevant as they did not relate to the 23 Common Issues which Mr
Page 499 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Justice Fraser had to decide.®”” Upon reading the Common Issues Judgment, I
found it difficult to understand why Mr Justice Fraser had made the extensive
factual findings he did in respect of Mrs Stockdale’s case (often in terms implicitly
or explicitly critical of POL), yet had declined to make any findings about whether
she had knowingly submitted inaccurate accounts which had the effect of
obscuring the shortfalls about which she complained. To be clear, POL’s position
had been that Mr Justice Fraser should not make findings as to credit based on
Mrs Stockdale’s post-contractual conduct at all (including any evidence as to
false accounting), since this was not relevant to the Common Issues and would
impinge on future trials — see further §768 above and §981 below. However,
given that he had considered a number of post-contractual matters, I could not
immediately see why he had declined to address the evidence of false
accounting in Mrs Stockdale’s case, instead holding that “[wJhether she was right
to act as she did at the time regarding her accounts is a matter for another trial”.
920. Third, I am asked to explain the basis for describing the Judge’s approach as
“astonishing; it is unfair and unprecedented” in paragraph 1.6 (which went on to
give a preliminary recommendation that an appeal be lodged). In particular, I am
asked to describe any conversations that I had with Counsel on the matter.
921. My email records show that I sent a draft of the Summary Note to David
Cavender QC and Gideon Cohen (who were the Counsel team for the CIT) on
the evening of 9 March 2019 for their review.®’6 In that email, I described it as a
“very early and initial summary of the key findings in the Judgment’, and
specifically drew their attention to paragraph 1.6. I expressed my hope that they
877 WBONO0001366.
878 WBON0001463.
Page 500 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
would agree with that paragraph but asked them to “say if [they didn’t/’. Gideon
Cohen responded with some non-substantive amendments.®”° David Cavender
QC responded on 9 March 2019 that he had “no hesitation in agreeing with [my]
clause 1.6."88° As I recall, I had discussed the substance of paragraph 1.6 point
with David Cavender QC on several occasions after we received the draft
Common Issues Judgment (and thus, the email chain referred to above does not
represent the totality of discussions we had on this matter, though it reflects their
broad tenor). As I understood it, David Cavender QC was of the view that Mr
Justice Fraser had gone badly wrong in making so many factual findings which
went to matters of performance and breach.
922. Both Counsel and I felt that Mr Justice Fraser’s approach was “unfair’ in this
regard, as (i) he had made those factual findings even though POL had not yet
presented full evidence on these matters, and (ii) they were properly matters for
other trials which still had to be held. We were not reassured by Mr Justice
Fraser's occasional statements in the Common Issues Judgment that he was not
making findings on matters of performance and breach, as it seemed to us clearly
arguable that, on a fair reading of the judgment, he was. For example, at [558]
he made critical observations about the quality of the NBSC helpline (“on the
evidence before me, the Helpline did not operate for the Lead Claimants in the
manner that the Post Office contended for ...”) and went on to temper this by
saying, “detailed findings of fact as to this must however wait for a later trial’
(emphasis added).**' Similarly, at [955], Mr Justice Fraser described POL’s
system of training for incoming SPMs as “contrary to business logic” and found
878 WBONO000650.
880 WBON0000649.
881 See to similar effect, [569(57)] of the Common Issues Judgment.
Page 501 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
that “the subjective experiences of the Lead Claimants so far as training was
concerned was far from ideal’. At the time, the fact that he had (at [954]) prefaced
these observations by saying that they were “non-binding” did not seem to me or
David Cavender QC to be of much comfort.
923. Further, both Counsel and I considered that the nature and range of terms which
Mr Justice Fraser had implied into the contracts between POL and SPMs - either
as incidents of the duty of good faith which he implied, or as freestanding terms
implied on grounds of business efficacy — to be “unprecedented” bearing in mind
the state of the law on this.
Advice from David Cavender QC on appeal and recusal (Q107 to Q108)
924. Having indicated his agreement with paragraph 1.6 of the Summary Note, David
Cavender QC’s email of 9 March 2019 then floated the possibility of applying for
Mr Justice Fraser to recuse himself as Managing Judge, as follows:
“Indeed I am beginning to form the view that seeking the Judge's recusal
is something that we need to actively consider. I am drafting a Note too
(as requested) dealing with the appeal on a high level. As part of that I
have been collating some of the Judges comments and findings and the
gross procedural irregularity here. If we are right about all of those points
- then how could a reasonable independent observer think that such a
Judge could fairly adjudicate in the future on a dispute between the
parties i.e. trial 3 and beyond? I recognise its extreme - and being so
directly involved makes it difficult to be objective - but what we have been
served up with is frankly so shockingly bad that we must at least consider
it. I deal with this point — in brief in my Note which you will get tomorrow.”
925. With reference to Q108 of the Request, this was (to the best of my knowledge
and recollection) the first time the possibility of seeking Mr Justice Fraser's
recusal was raised with me.
Page 502 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
926. The following day, 10 March 2019, David Cavender QC provided a note of advice
setting out his preliminary views on the prospects of appealing the Common
Issues Judgment.®®2 In short, he recommended an appeal and suggested that it
should be focused on: (i) Mr Justice Fraser’s “errors in construing the SPMC and
NTC contracts, the terms to be implied into them, and the relationship that arose
as a result’ by reference to “what happened post contract’ (paragraphs 10 and
27); and (ii) the “gross procedural unfairness exhibited by his making findings of
fact on unnecessary matters based on partial information” (paragraphs 10 and
12 ff). In that note, he elaborated on his suggestion that POL should consider
applying to remove Mr Justice Fraser as the Managing Judge. He observed that:
15.1 The grounds for recusal were essentially the same as those on which the
‘gross procedural unfairness’ limb of the appeal would (if brought) be
based (paragraphs 12-18).
15.2 There “is a very high threshold to justify such an application. This is
undoubtedly the nuclear option — but this Judgment is very bad indeed.
The way [Mr Justice Fraser] has conducted himself in this matter in my
view is unjudicial and is unprecedented” (paragraph 11).
15.3 The fact that there were a number of further trials to come over which Mr
Justice Fraser would preside was a “special and odd feature” of the case
which made the problem of his possible or apparent bias more acute
(paragraph 20).
882 WWBON0001466; POL00267565. This was prepared at POL'’s request as part of its mitigation
planning for Monday 11 March 2019, see: WBON0001446; POL00267481. I sent David
Cavender QC’s note of advice to POL along with the Summary Note on 10 March 2019:
WBONO0000205.
Page 503 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
15.4 POL should consider “instructing separate Counsel to consider this point
on recusal — as having been so involved in this it is difficult to be truly
objective”(paragraph 23).
927. With reference to Q107, my best recollection is that I only first considered the
possibility of recusing Mr Justice Fraser when it was raised by David Cavender
QC and only appreciated that there were arguable grounds for this upon reading
his advice note. I had no prior experience of recusal applications and no
knowledge of the law on bias, so I was guided by Counsel's views on these
matters. I recall thinking that it was a sound idea to get a second opinion from
separate Counsel. I could also understand why David Cavender QC had raised
the possibility of recusal given the general tenor of the Common Issues
Judgment, and the fact that to my mind, Mr Justice Fraser appeared to have
prejudged a number of factual matters that would be in dispute in later trials.
(ii) Strategy and criticism following the Common Issues Judgment (Q105 to
Q106)
928. Q105 of the Request asks me about my reflections on POL's litigation strategy in
light of the Common Issues Judgment. Prior to the Common Issues Judgment
being handed down, POL had already taken steps to modify its litigation strategy.
As explained above at §§385 ff, Mr Justice Fraser had earlier criticised both
parties at the October 2017 CMC for adopting what he felt was an insufficiently
cooperative approach to the litigation. At that stage we had commissioned David
Cavender QC’s ‘Five Things’ review; sought to tone down correspondence with
Freeths and invited them to jointly ‘reset’ the relationship with us; and POL sought
to limit the number of applications it made. Although I sensed that the September
Page 504 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
2018 application to strike out parts of the Claimants’ evidence was perceived by
Mr Justice Fraser as an excessive step, I have explained above (at §§697 ff) the
background to that application and why it was considered to be necessary.
929. At the time the draft Common Issues Judgment was circulated, the strategy for
the HIT had already been set in motion and my focus was on this trial as it was
due to begin the next working day. I do however recall having some discussions
with Tony Robinson QC to the effect that we should somewhat soften the tone
that we adopted in the HIT; in particular, we discussed the importance of treading
lightly in submissions and cross-examination. We did not want to receive further
criticism for being heavy-handed. That said, Counsel had to balance this against
the fact that he still needed to cross-examine witnesses on some difficult points,
including putting to the SPM witnesses that the problems they encountered in
their branches may have been caused by their own actions, and not by Horizon.
930. Beyond this, decisions on strategy in terms of appealing the Common Issues
Judgment and making the Recusal Application were strongly guided by advice
from Counsel. For example, I recall that Mr Justice Fraser's approach to implying
terms into POL’s contracts with SPMs featured in discussions as to the approach
to be taken to appealing the Common Issues Judgment, and in particular I recall
that the POL Board Subcommittee were minded to adapt POL’s case on appeal
to make some concessions in light of that approach. I recall that Lord Neuberger
and Lord Grabiner QC (whose involvement I deal with further below) advised
against that course, as making those concessions would undermine the
coherence and clarity of POL’s position on the law. They were also of the opinion
that making such concessions was a slippery slope and could lead to further
Page 505 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
terms being implied that could be damaging to POL’s business in the long run;
see further below, §§971-975.
931. As to Q106, I do not recall any representative of POL directly challenging me on
WBD’s conduct of the group litigation after the Common Issues Judgment was
handed down, although there was understandable concern that POL did not
appear to be faring well and about what implications this may have for the
business. At this time, Paula Vennells had recently stepped down as CEO and
Alisdair Cameron had assumed her role. At around this time, I became aware
that POL had brought in Norton Rose Fulbright and subsequently (following Mr
Justice Fraser's Recusal Judgment) HSF. Although this is speculation, at the time
I considered that this move probably flowed from a loss of confidence in WBD or
at least a desire to have a fresh perspective. It sometimes happens that a client
changes solicitors or counsel, or both, after a setback in a piece of litigation.
Norton Rose did not express the view to me that WBD’s conduct of the litigation
to date had been inappropriate (nor did HSF when they took over from Norton
Rose). Later, in May 2019, David Cavender QC was replaced by Helen Davies
QC for the appeal against the Common Issues Judgment on the advice of HSF
(I was not consulted on that decision and was only told about it after the decision
was made).
(iii) Further advice on recusal — instruction of Lord Neuberger and Lord Grabiner
QC, POL Board meeting on 18 March 2019 (Q104, Q109 to Q112)
Instruction of Lord Neuberger and Lord Grabiner QC (Q109)
932. Returning to the chronology of the Recusal Application, on Monday 11 March
2019 there was a meeting of the POL Group Executive to consider the draft
Page 506 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Common Issues Judgment. Given that I was fully occupied with the first day of
the HIT, Amy Prime attended on my behalf to take a note and report back.
Following the meeting she sent me and Tom Beezer a note of what was
discussed.® It contained the following note of discussions on the subject of
recusal:
“On the recusal decision, Al [i.e. Alisdair Cameron] was keen to press
ahead with recusal now, during the Horizon Trial however Jane / David
explained how bad this could go and this idea seemed to be dropped but
we would get another opinion on this from a super silk.”
933. More generally POL wanted a senior silk to review the Common Issues Judgment
and advise on the prospective appeal and the strategy for appealing. It was
typical of POL to seek a further review by new lawyer in this sort of situation (as
had happened when advice was sought from Linklaters during the Mediation
Scheme, by Jonathan Swift QC after the scheme, and by David Cavender QC
when he was brought in and asked to conduct his ‘Black Hat Review’), so this
did not come as a surprise to me. Amy made enquiries of One Essex Court and
Lord Neuberger and Lord Grabiner QC were put forward by the clerks as possible
options for the ‘super silk’.®°* On the evening of 11 March 2019 WBD put both
options to POL with a suggestion that Lord Grabiner QC be retained, since he
would be able to represent POL in court (which Lord Neuberger could not, being
a retired judge).®*5 The following morning Jane MacLeod expressed the view that
Lord Neuberger would have “insight into the thinking of the Court of Appeal
judges and their approach to current trends of contractual interpretation” and
883 WWBONO000652.
884 WBONO0000653
885 WBONO000654
Page 507 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
asked to discuss further with Tom Beezer. I cannot specifically recall what I
thought of this at the time, but can see from my email records that I sent Tom a
note sounding a (slight) note of caution in the following terms:8®>
“Remember Neuberger only around for a couple of days.
Also, he wrote the law on interpretation of contracts — orthodox legal
principles — so has skin in the game because this judgment crashes
through his doctrine on contract interpretation. Will he be open minded?”
934. I also informed Tom that David Cavender QC had initially suggested that POL
instruct both Lord Neuberger and Lord Grabiner QC, but had then moved away
from Lord Neuberger as he had limited availability and would not be able to be
the advocate for the Recusal Application “which should happen asap if approved
/has merits” (although I do not specifically recall this particular conversation with
David Cavender QC). On a separate email chain at around the same time, Tony
Robinson QC suggested Lord Pannick QC as a ‘super silk’ who would be
interested in the recusal issues.®°” (My recollection is that Tony Robinson QC
had very little input into the Recusal Application, due to being fully immersed in
the HIT). I therefore contacted Blackstone Chambers to make enquiries but was
informed that Lord Pannick QC was not available.®® I do not recall, and my email
records do not suggest, that I had any other correspondence with clerks or
alternative counsel prior to the decision to instruct Lord Neuberger being made
by POL (which it was later that morning).°®° POL’s ultimate decision was (i) to
instruct Lord Neuberger to advise on “the merits of a recusal application and
whether such an application should be made” as well as the prospects of
886 WWBONO000655.
887 WBONO000656.
888 WWBONO000657.
889 WBONO000658.
Page 508 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
appealing the Common Issues Judgment, and (ii) to reserve Lord Grabiner QC:
POL00023930 (email sent by Amy Prime email to the One Essex Court clerks
on 12 March 2019). The latter step was taken because Lord Neuberger could not
appear in court on POL’s behalf, so Lord Grabiner QC was reserved in case Lord
Neuberger advised POL to go ahead with the Recusal Application. As Lord
Grabiner QC’s availability was limited, since he was due to go under brief from
30 April 2019, enquiries were also made of Mark Howard QC at Brick Court
Chambers, although I was not directly involved in making these enquiries.8°°
935. Lord Neuberger was provided (by WBD) with five bundles of papers which
included the transcripts of the CIT (cf. Q109.4).®°' He was also briefed orally by
David Cavender QC,*° and was provided with a note of the background to the
recusal application prepared by David Cavender QC and Gideon Cohen (though
I do not believe WBD had sight of this background note until it was provided along
with Lord Neuberger’s note of advice on 14 March 2019).®° I did not have any
direct contact with Lord Neuberger before he produced his note of advice, and
no indirect contact save for the emails with his clerks that I was copied into.
David Cavender QC was in touch with him as I have set out above, but I was not
aware of the detailed content of their conversations (cf. Q109.1).
Lord Neuberger’s note of advice (Q110)
890 WBON0000658; POL00023988; WBON0000660. Tom subsequently stood Mark Howard QC
down in respect of the recusal aspect once POL decided to formally instruct Lord Grabiner QC
(see below, §937): WBON0000665.
891 WBON0001468; WBONO0001469; cf. WBON0001470.
892 WBON0000659.
893 WBON0001474. The Counsel team’s background note is: POL00371317.
894 My diary indicates that I attended a meeting with David Cavender QC and Jane MacLeod on
14 March 2019 following receipt of Lord Neuberger’s note of advice, though I cannot recall the
specifics of that meeting and my firm's records do not suggest that Lord Neuberger was on that
call.
Page 509 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
936. By Q110 I am asked to explain what my views were of Lord Neuberger’s note of
advice on the recusal aspect dated 14 March 2019 (POL00025910). In summary:
936.1.First, Lord Neuberger indicated that he had “looked at the Judge's
reasoning and conclusions on the interpretation issue” albeit “only very
cursorily’ (as his fuller advice on that aspect was to follow in due course).
His provisional view was that there were some “quite significant points on
which the PO has a reasonable case, and, at least on the face of it, some
points on which the PO has a pretty strong case.” In his view, Mr Justice
Fraser's “relational contract’/good faith justification for the implication of
terms ... is controversial in itself, but, quite apart from that, [he] appears to
have extended its application in a fairly radical way’; further, Lord
Neuberger noted that “there is often very little or no reasoning offered to
justify the implication of terms on the alternative, and conventional, basis
of necessity” (paragraph 5). Overall, he commented that he was “left with
the uneasy feeling that the real justification in the Judge's mind for the
implication for at least many of the terms which the Judge implied was the
raft of adverse factual findings that he has made” (paragraph 6). These
outline and preliminary thoughts broadly accorded with my own view that
Mr Justice Fraser had likely erred in his approach to the Common Issues.
936.2.Second, Lord Neuberger observed that there was “real force” in the
argument that Mr Justice Fraser had acted unfairly in “[making] the
findings about the factual evidence and the witnesses that he did”,
particularly given that POL had explained why it had not led detailed
evidence on many of those issues (paragraph 7). Again, this broadly
Page 510 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
accorded with my own views of that Mr Justice Fraser’s approach in the
Common Issues Judgment.
936.3. Third, in relation to the recusal issue, Lord Neuberger was “struck by the
fact that ... many of the paragraphs in the Judgment are given over to
descriptions of evidence, and findings of fact, in relation to what happened
after the contracts had been entered into, often in trenchant, even highly
critical, terms. And, importantly, as I understand it, those descriptions and
findings relate to witnesses who will be called at later trials and evidence
which will have to be considered at later trials” (paragraph 8). This
accurately reflected my understanding of the factual basis of the proposed
recusal application.
936.4.Fourth, Lord Neuberger concluded that these matters gave rise to
“reasonable grounds for PO to bring an application to recuse the Judge in
these proceedings” (paragraph 19). He identified that the Court of Appeal
might take the view that Mr Justice Fraser was entitled to deal with the
evidence before him; that his findings would not impinge on future trials;
and/or that he had “gone out of his way to make it clear that he was not
making conclusive findings”. Ultimately however he did not think that these
were serious obstacles to POL’s case on recusal in the circumstances
(paragraphs 12 to 18). As to this, as noted above I did not have any
knowledge of the law of bias at this time. I had never before had occasion
to make a recusal application (which I understood to be exceptional and
unusual) and so was reliant on Counsel to advise on this point. However,
Lord Neuberger’s advice was reasoned and logical, aligned with my
understanding of the litigation and how it had developed, and overall it
Page 511 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
supported my view of how and where Mr Justice Fraser had overstepped
the mark in the findings he had made in the Common Issues Judgment.
936.5.Fifth, Lord Neuberger advised that the recusal application should be made
(if it were to be made) as soon as possible after the Common Issues
Judgment was formally handed down, which would likely mean making the
application without notice and during the HIT: “The fact that this course
would be taken without notice and after the present trial has begun cannot
be blamed on the PO: until they have the Judgment, they are not ina
position to take a view on the recusal issue” (paragraphs 20 to 22). Again,
this recommendation made logical sense to me although I had no doubt
(as Lord Neuberger recognised) that this would be a tricky and
controversial application to bring in the midst of the HIT.
Recusal Paper (Q104)
937. Against this background, on 15 March 2019 Jane MacLeod emailed Tom Beezer
and me forwarding an email she had sent to Tim Parker (POL’s Chairman) and
Tom Cooper (UKGI’s representative on the POL Board), outlining Lord
Neuberger’s advice and setting out her proposed next steps (including her
intention to formally brief Lord Grabiner QC so that he could start reading in).6%
She advised us that POL were setting up a Board call for 5pm on Monday 18
March 2019, for which a “plain English paper’ would be needed addressing the
following matters:
“Why we are considering a recusal application
What the application (if successful) will achieve
895 POL00023898.
Page 512 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Risks of not proceeding
Prospects of success: what advice have we received, who from (LNQC
but given speed with which it was produced - is it fully considered?; Will
LGQC have read in sufficiently by then to also be able to offer an
opinion? ‘why we should believe them?’)
risks
Process & timing”.
938. The paper that was ultimately prepared was POL00023955 (the “Recusal
Paper’). Its aim was therefore to summarise the background to the proposed
Recusal Application, the relevant legal issues, and the advice received, in a way
that was clear and accessible to non-lawyer Board members. I do not specifically
recall the Recusal Paper, however my firm's records indicate that it was a
collaborative effort between myself, Tom Beezer, Jane MacLeod and David
Cavender QC; and approved by Lord Grabiner QC. At this stage, my focus was
on the ongoing HIT, and therefore Tom Beezer took the lead on producing the
Recusal Paper.
939. Tom produced a first draft on 15 March 2019 which he then circulated to me,
David Cavender QC, Gideon Cohen, and Amy Prime for comment.®® I redrafted
the Recusal Paper later that evening; I did not include my own views on the
prospects of success (since Counsel were leading on this), but recorded Lord
Neuberger's key conclusions and left a placeholder for David Cavender QC to
insert his views on the merits. In my covering email, I queried whether Jane
MacLeod had asked us to make a recommendation to the Board; Tom indicated
that he would raise this with Jane MacLeod, subject to which his view was that
896 WWBON0001493; WBONO001494.
897 WBON0001495; WBON0001496: “DAVID TO ADD HIS VIEWS ON RECUSAL HERE IN HIS
OWN WORDS’.
Page 513 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
we should simply set out the issues and let the Board decide.®% I can see from
my firm’s records that Tom sent the draft as it stood to Jane McLeod on the
morning of 16 March 2019 (on the understanding that further amendments would
be made to it to that day).®®° It appears that this was because Jane MacLeod had
a call with Tim Parker, Alisdair Cameron and Kelly Tolhurst MP at 12.15pm that
day.
940. Shortly after Tom sent the draft on, David Cavender QC circulated a further
version of the Recusal Paper with his own changes to my draft in track, including
several paragraphs setting out his views on the merits. He felt that, as legal
advisers, we ought to be setting out a recommendation and to that end his
revised draft set out that: “Therefore, Mr Cavender’s view is that it is difficult to
see a realistic alternative and so a recusal application should be made.”
941. Later the same day I responded with my thoughts on David Cavender QC’s
revised draft, to the following effect:
“L agree with David about offering a recommendation and would support
his recommendation.
There is now an almost constant doubt hanging over how to conduct the
Horizon litigation because we feel that we are always a whisker away
from the Judge attacking us.
I think we should add a point to the risk section that making the
application may reinforce the ‘arrogance’ attack on PO but that attack
has already been made and it may well be made again, so there is no
guarantee that staying quiet will protect POs brand from repeat attacks.
Also, I think we should add in the timing section that a recusal application
might encourage the C of A to move quicker on the main appeal. These
898 WWBON0001497
888 POL00022960.
800 WBON0001499; WBONO001500.
Page 514 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
issues are all interconnected and the C of A is unlikely to want to leave
a recusal application hanging over the litigation. David, do you agree?
And also that the above point on timing may limit the amount of
operational change PO needs to undertake in the short term to comply
with the judgment, which may be wasted cost if the judgment is
overturned on appeal.’
942. Amy Prime then worked up a new draft to implement David Cavender QC’s and
my comments.°? Inter alia, the Recusal Paper now included the following
concluding “Recommendation”:
“Recommendation
Although a recusal application is difficult and comes with substantial
risks, for the reasons stated above, both Mr Cavender and Womble Bond
Dickinson recommend that the application is made as soon as possible.”
943. Tom sent clean and compare versions of the updated draft to Jane MacLeod later
that day, i.e. 16 March 2019.%° It was still somewhat in draft form, and contained
a single comment from myself about a particular paragraph in the draft (my
comment is highlighted in yellow):
“Aside from the above legal points, we would also note that several of
Post Office's witnesses, many of whom are long serving employees,
were good enough to give evidence in Court for Post Office and have
now had their reputations tarnished. It is of course a matter for Post
Office to determine the extent to which it now wishes to try to protect its
staff from criticism. [Jane = this point may be better made verbally so we
can remove it.]”
944. My firm’s email records show that it was Tom and Jane McLeod who took the
Recusal Paper forward (though I remained either an addressee of these emails
or in copy, and I would have spoken to Tom on the phone from time to time about
it). It appears that Tom and Jane MacLeod had a call to discuss the draft paper
8°1 WBONO000666.
802 WBONO0000667; WBONO000668.
8°3 POL0002391 1; POL00268458; POL00268459.
Page 515 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
on 16 March 2019,2% following which Jane McLeod made changes including
restructuring the draft.°°> Further, she pulled together some of what she viewed
as the more “egregious comments” made by Mr Justice Fraser that might
demonstrate bias, to be placed in an Appendix to the Recusal Paper.°° Tom
circulated a further version of the Recusal Paper, incorporating and responding
to Jane McLeod’s comments, on 17 March 2019.2” The draft still contained some
highlighted comments requiring resolution; these required input from Lord
Grabiner QC. Jane McLeod made a few minor suggestions, which she otherwise
approved to be sent to Lord Grabiner QC.°8 Tom then finalised the Recusal
Paper and sent it to Lord Grabiner QC’s clerks.9°9
945. Lord Grabiner QC approved the content of the Recusal Paper on the evening of
17 March 2019 (“Treat this as my broad ‘yes that is ok’.”).9° Tom therefore
removed the highlighting which indicated that we were awaiting Lord Grabiner
QC’s approval and sent a clean version onto Ms McLeod (this is the final version
of the Recusal Paper that the Inquiry cites, POL00023955).
946. My involvement in the drafting process as set out above indicates that I broadly
agreed with the content of the Recusal Paper. In particular, I note that I supported
the inclusion of an express recommendation to the POL Board, and expressed
904 POL00330036.
8°5 POL00023231; POL00268479.
906 POL00023229; POL00268503.
8°7 POL00022969; POL00268516.
808 WWBON0000672.
88 WWBON0000673; WBON0001501. The version of the Recusal Paper as sent is: POL00268533.
Earlier that day Tom had emailed the clerks at One Essex Court, noting that POL sought to
understand whether Lord Grabiner QC agreed with the proposed Recusal Application, and if
so, that Jane MacLeod wished to have a call with him the following day prior to the planned
Board meeting (with Lord Neuberger to attend the meeting itself): WBON0000671. Lord
Grabiner QC had responded to that email indicating that he was in “broad agreement’ with Lord
Neubeger's position (though I was not in copy and received this email as part of a later chain):
WBONO001501.
10 WBONO000675.
Page 516 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
my agreement with David Cavender QC in this respect, viz. that the Recusal
Application should be made. As I have said, however, my view was primarily that
the application should be made as a matter of strategy; my knowledge of the law
on recusal was limited and I was naturally deferring to Counsel’s expertise as
regards the substance of the application. With reference to Q104.2 and Q104.3,
my “concerns as to Fraser J in respect of bias” at this point in time can therefore
be summarised as follows:
946.1.1 was concerned, first, that Mr Justice Fraser had formed a negative
opinion of POL based on incomplete evidence. At the CIT, POL had not
led evidence on post-contractual matters®'' and had therefore not fully
responded, in respect of each Lead Case, on such matters such as what
training and support had been provided to each Lead Claimant; the exact
chronology of their accounting records; what investigations had been
undertaken and why; and so on. My view was that once complete
evidential picture was available, a judge may well have formed different
conclusions on the overall responsibility for losses than those to which Mr
Justice Fraser had appeared to reach. This is not to say that matters would
have been resolved entirely in POL’s favour, but I believed that there was
a good chance of a more balanced determination. I therefore felt that Mr
Justice Fraser had acted prematurely in making some of the observations
that he did. This was especially so given that we had made clear to him in
closing that the scope of the evidence tendered by POL was limited; what
°11 Save for a few narrow areas where post-contractual evidence was required — see for example
paragraph 33 of the witness statement of John Breeden, to which I referred above at fn. 758:
WBONO001351.
Page 517 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
the reasons for this were; and that it would therefore be unfair to make
factual findings on post-contractual matters at this stage.
946.2.Second, and more specifically, I recall that my reading of the Common
Issues Judgment at the time was that the Judge had already begun to form
his view on specific issues around Horizon and POL’s accounting
practices. There were, as I read them, findings of fact on these points
within the Common Issues Judgment. I struggled to see how the Judge
could change his mind on these issues in later trials where those points
would be explored in more detail. This caused me to believe that he may
have (or appeared to have) closed his mind to key points that would form
the focus of those later trials.
946.3.Third, as I have explained above, although Mr Justice Fraser had stated
within the Common Issues Judgment that he was not making findings on
issues of contractual performance and breach, my reading of the judgment
taken as a whole was that he had made findings of fact that were central
to those issues. Lord Neuberger’s advice dated 14 March 2019 had
affirmed this view, in that he had explained that the fact that a judge asserts
that they are not making particular findings is not determinative:
“In my view, the Judge's attempts to distance himself from, or to
water down, his illegitimate findings, in some ways render them
worse rather than better. What was he doing making findings
(sometimes in trenchant, even damning terms about the PO's
witnesses, and exculpatory or better about several of the
Claimants), if he knew that the findings were, at best,
unnecessary, indeed inappropriate?” (paragraph 18).
Page 518 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
946.4.Fourth, I was concerned that Mr Justice Fraser had been highly critical of
POL's witnesses including Angela Van Den Bogerd. I have previously set
out why I felt this criticism was unfair in response to the Inquiry’s Q92
above (§754). In brief, I felt that it was unfair to criticise Angela Van Den
Bogerd for not having told the “full story” in her statement, when she was
advised not to do so because this would have required her to give
evidence on matters POL’s legal team considered were irrelevant to the
Common Issues. Given that Angela Van Den Bogerd was one of POL’s
factual witnesses in the HIT, I was concerned that Mr Justice Fraser would
carry forward his view of her, which I felt was unfairly formed, into his
assessment of the evidence in that trial.
946.5.Fifth, however, I repeat that these were my views on the matters which
formed the factual basis for the Recusal Application. As to the application
of the law of bias to these matters (including the “fair minded and informed
observer’ test), I was guided by the views of Counsel and in particular,
David Cavender QC, Lord Neuberger and Lord Grabiner QC, all of whom
had advised that there was a sound basis for making the application.
947. As to whether representatives of POL told me of their concerns in respect of bias
(cf. Q104.2), I recall that Jane MacLeod and Rodric Williams held similar views
to the above. I do not recall having communicated directly with other
representatives of POL on these matters.
POL Board meeting on 18 March 2019 and subsequent decision-makin: 111
to Q112)
Page 519 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
948. Prior to the Board meeting on 18 March 2019, Jane MacLeod attended a
conference with Lord Grabiner QC. David Cavender QC and Gideon Cohen also
attended, as did Tom Beezer for WBD (I did not attend as this was then the fifth
day of the Horizon Issues Trial). WBD’s file note for this conference sets out that
Lord Grabiner QC’s advice was as follows:
“Procedural Structure: ... Lord Grabiner commented that the Judge
had "trespassed onto matters that are for later trials" and the
fundamental problem that builds into the sequential trial structure is that
those findings and opinions (which should not have been made or
voiced) will be carried through into the later trials and that is the
"perceived bias" that gives rise to the need for a recusal
application. Lord Grabiner commented that the case management
displayed in this matter was extremely poor. It was noted that the
problems now experienced by Post Office and manifested in the CIT
Judgment were predicted multiple times before this Judge at the making
of the GLO and many times after. Lord Grabiner also noted that Post
Office had attempted to deal with the issue by applying for the striking
out of the Claimants evidence that was irrelevant to the CIT, yet the
Judge had refused to do that. The Judge had sufficient warning of the
risk of him taking into account evidence that was irrelevant to the CIT
and taking into account post contractual matters in a trial supposed to
be confined to construction issues only; however he had "not been able
to restrain himself". As an aside, the number of implied terms found by
the Judge was wholly extraordinary but that was to be a matter for an
appeal on law — but Lord Grabiner expressed his shock that an English
High Court Judge could have arrived at the CIT Judgment.
Urgency: An application for recusal should be made urgently ...
Duty to act: ... An appeal on the law may correct some of the very
significant errors in the CIT Judgment but then the case will be sent back
to this Judge who has demonstrable apparent bias against Post Office
and hence the firm conclusion that Post Office will lose and the financial
impact of that will be substantial. Recusal is therefore essential and Lord
Grabiner asserted that in the face of legal advice from Lord Neuberger
that recusal should be applied for and the quantum of damages that Post
Office will pay out on a loss, then it was Lord Grabiner's view that there
was a duty on Post Office to seek recusal. Lord Grabiner stated that in
Page 520 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
his view the Board of Post Office had no option but to seek recusal"
(emphasis in original).9'2
949. The file note records that Lord Grabiner QC’s conclusions were that: (i) “there
are strong arguments in favour of an application for recusal’; (ii) it was his “strong
view that a recusal application was the right course of action’; (iii) there “[was] a
‘serious prospect of success”, and (iv) Mr Justice Fraser had, in his view "done
‘an unbelievable nonsense and demonstrated apparent bias”.
950. Later that day the POL Board meeting took place, the minutes of which are
POL00021562 (to which I am referred by Q111 of the Request). I did not attend
that meeting and the minutes do not indicate that anyone else from WBD did
either. I do not recognise the minutes and I believe that the first time I received
them was with the Inquiry’s Request. This is confirmed by the absence of the
minutes from my firm’s file.
951. Following the Board meeting, on 18 March 2019 Jane MacLeod sent me, Tom
Beezer, and Rodric Williams an email update in the following terms:
“[Lord Neuberger] was very balanced in his approach, but confirmed that
he thinks we have a good case on recusal. The Board asked a number
of questions and my sense was that they were ‘calmed’ by his
discussion. However they haven't yet made a decision. There is a further
board call on Wednesday [20 March 2019] at 12.30 and they have
requested whether Lord Grabiner would be available in person at the
time — ideally at FD if that’s possible? Having said that they recognise
that he will almost certainly say the same things as DNQC.
912 POLO0268834. My email records indicate that I received this file note on 20 March 2018:
POL00022883. Later that day Tom Beezer send Jane MacLeod a slightly amended version of
the note, together with a comment from Gideon Owen that he had “no corrections or changes
to make”: POL00330038. The amended note was not materially different from that quoted
above.
Page 521 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
[ul
There is significant pressure to be able to say how we are going to treat
those claimants who establish they have a case, and all those outside the
scheme who may have a similar fact pattern. So what they have asked
for is a pro forma model of what the various outcomes could be — that is,
what is the financial impact such that taking a step like recusal is ‘worth
it’.”913
952. On the morning of 20 March 2019, Jane MacLeod emailed Tom Beezer (with me
in copy) querying whether an alternative approach might be viable, involving: (i)
applying to Mr Justice Fraser to adjourn the ongoing HIT on the basis that POL
was seeking to appeal the Common Issues Judgment on procedural fairness
grounds (and that if that appeal was successful, “it would also put the fairness of
the Horizon trial at risk’); (ii) if an adjournment was refused by Mr Justice Fraser,
“[seeking an] order from a higher court to the same effect’, or alternatively
seeking recusal at that stage if no such remedy ‘existed in law’.9'4 From the
framing of Jane MacLeod’s email, I infer that this query was raised by a member
of the POL Board in anticipation of the call later that day.
953. This proposal was put to David Cavender QC, Gideon Cohen and Stephanie
Wood (who was by then instructed to assist with the Recusal Application). I
expressed the view that the proposal did not make much sense since the
‘procedural unfairness’ reflected in the Common Issues Judgment did not in and
of itself ‘infect’ the HIT, rather, the ‘cross-infection’ occurred if Mr Justice Fraser
was biased or had the appearance of being biased. In any event, I thought that
Mr Justice Fraser would be very unlikely to adjourn the HIT on the basis
°13 WBON0000677. My email records indicate that I did not respond to that email, although Rodric
Williams and Tom Beezer added some observations: WBON0001511.
14 WBONO001714.
Page 522 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
suggested in Jane MacLeod’s email.°° David Cavender QC responded
essentially agreeing with my reasoning and adding some further points as to why
the proposal was misjudged:
“I would advise strongly against the proposed course. I say this for the
following reasons:
The immediate (and likely irreversible) prejudice PO are suffering is the
effects of the apparent bias Fraser J showed in the CIT upon his current
handling of the Horizon issues trial. There is also the future prejudice of
him handling the breach trial in November 2019.
The only way of seeking to deal with the prejudice is to seek his recusal
on an urgent basis.
Seeking to appear before him indicating that PO is going to appeal
against his CIT judgment on grounds of procedural unfairness — will
assuredly not result in him adjourning the Horizon trial. He will not do so
because the unfairness in the CIT trial itself does not infect the Horizon
trial. It is the apparent bias of Fraser J that infects the Horizon trial. The
only remedy for that is recusal.
If, on this proposal, Fraser J’s refusal to adjourn the Horizon trial is then
appealed to the Court of Appeal — they would assuredly not adjourn that
trial and would not recuse him- because there would not application
before them to do so.
Furthermore, an appeal against a refusal to recuse is much more likely
to come on as an urgent appeal — than an appeal against the refusal of
a judge to adjourn a trial on the basis that he showed procedural
unfairness in an earlier trial between the same parties. Indeed the latter
appeal is very likely to come on after the Horizon trial is completed and
the Judgment handed down. This fact would make it more unlikely the
Court of Appeal would intervene.
And, if all this comes to pass (as it most assuredly would) is the proposal
that then PO applies to the judge to recuse himself ? And then appeal
him if he does not ? This make no sense- and would all come too late to
be effective to deal with the prejudice in (1). Indeed, this course of action
would look very much as if PO were seeking to delay matters and
behave badly- in the manner presently charged by the Judge.
If there are good grounds for a recusal (and clearly there are) and good
prospects of success (as advised) then the Court of Appeal would expect
15 WBONO000679.
Page 523 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
PO to apply to the Judge to recuse himself and then appeal him if he did
not. There is no middle ground here” (emphasis in original).°"®
954. Tom forwarded this advice to Jane MacLeod ahead of the Board call (which had
by this time been moved to 11:45am).9'7 Jane MacLeod then raised further
questions on which the Board were likely to want assurance, namely (i) “if not
[the above proposal], then is there any other alternative to recusal”, and (ii) what
more would be achieved by recusal, if POL succeeded in appealing Mr Justice
Fraser’s interpretation of its contractual relationship with SPMs. Tom forwarded
these questions to the Counsel team including Lord Grabiner QC for
consideration in advance of the call.9"®
955. Again, I did not attend the Board call (this being Day 7 of the HIT); Tom Beezer
attended for WBD (alongside lawyers from Norton Rose) from 11:45am until
12:10pm. My firm's file note of the call records that Lord Grabiner QC outlined
the following advice to the Board:
"Lord Grabiner explained to the Board members that in his view the Judge
had been warned about admitting material into the CIT that should
properly be looked at only in later trials when proper evidence and
disclosure was before the Court. The Judge had rejected Post Office's
quite proper Strike Out Application and had appeared to appreciate what
the problem might be with inadmissible evidence from the Claimant
group, but at trial (the CIT) the Judge had gone well beyond his remit for
that CIT and made a range of findings on breach and the credibility of
Post Office witnesses. Lord Grabiner confirmed his view that the Judge
had behaved quite improperly and it was now right to ask him to stand
down. Lord Grabiner explained that it is apparent that this Judge has
concluded views on matters and as there are further trials to come, then
those concluded views would be a significant issue for Post Office unless
this Judge is asked to recuse himself.
816 WBONO0000681.
°17 WBONO0000682.
$18 WBON0000682.
Page 524 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Lord Grabiner confirmed that the apparent concluded views of this Judge
are so strong that there is no other way to deal with the issue than recusal.
If Post Office does not take such a step, yet later goes to the Court of
Appeal on matters of law from the CIT, the Court of Appeal will be left
wondering why such a step (i.e. recusal) was not taken.
Advice: ... Post Office has no option but to seek the recusal of this Judge
... whilst guarantees cannot be provided, Post Office does have a strong
case for recusal. Lord Grabiner confirmed that his strong
recommendation to the Post Office Board was to seek a recusal.”?"9
956. The file note further records that the POL Board asked a number of questions of
Lord Grabiner QC, as follows:
“A question was posed concerning the circumstance where recusal is
sought, but fails. ... Lord Grabiner pointed out that this Judge has
already formed a view of Post Office and so a failed recusal application
is unlikely to make a difference to outcomes when compared to plausible
outcomes from a situation where no recusal application is made.
A question was posed as to whether there was some "middle way" ?
Lord Grabiner explained that his firm view was that if Post Office does
anything short of applying for recusal, that strategy will fail. There is no
middle course which works and in that regard Lord Grabiner confirmed
that he had seen and agreed with an e-mail on the point send by David
Cavender QC.
A question was posed over a scenario where one assumes that the
criticisms of Post Office in the CIT Judgment were true ? Lord
Grabiner explained that in his view many of the implied terms and ‘good
faith' findings were wrong in law so if there were a different Judge then
the findings would be different as the assumed background and legal
background would be different so it is hard to place oneself in the
assumed position that all of the criticisms are correct. ... Theoretically it
is possible a different Judge could get to the same position — and if so
"so be it".
A question was posed as to whether Post Office could arrive at a
recuSsal scenario via a different route ? Lord Grabiner explained that
he could not think of any mechanism that would arrive at such an
outcome for Post Office, without Post Office making the application itself.
Lord Grabiner explained that if Post Office does not take the step to
apply for recusal then, as a certainty, it will have this Judge for all
19 POL00269796.
Page 525 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
following trials. Lord Grabiner explained that if he could have found
another way to proceed he would have discussed it with the Post Office
Board.
As a concluding comment, Lord Grabiner reminded the Post Office
Board that Lord Neuberger agreed with the advice on recusal that Post
Office was receiving and Lord Grabiner explained that there were few, if
any, more respected QCs and ex-Judges in this country”.9°
957. With reference to Q112 of the Request, Jane MacLeod emailed me, Rodric
Williams and others following the Board call (shortly after 2pm on 20 March 2019)
instructing WBD to issue the Recusal Application, with Lord Grabiner QC to be
instructed to undertake the advocacy."
(iv) The Recusal Application and Judgment (Q113 to Q115)
Preparation for the Recusal Application (Q113.
958. WBD recommended, and POL agreed, that work should start on the documents
necessary to support the Recusal Application on a provisional basis in advance
of the Board making a decision (i.e. on or around 15 March 2019), so that there
would be minimal delay in issuing the application.°22 As noted above, Lord
Grabiner QC was formally instructed at the same time so that he could start
reading in.°2° The overall view was that the application should be issued with
minimum delay if the Board gave its approval. It was anticipated that if this was
done by Friday 22 March 2019, the application itself might then be heard in the
week beginning 25 March 2018, during which the HIT was not scheduled to sit in
any event.924
820 POL00269796.
21 WBON0000683.
822 Cf. WBON0000661
£23 POL00023898.
924 See for example: POL00167515.
Page 526 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
959. Substantive preparation for the Recusal Application was largely undertaken by
Counsel; I recall that the Counsel team, and Lord Grabiner QC in particular, had
a clear idea as to how they wished to run the application. WBD’s main
contribution was to the witness statement in support of the application. Work
began on a witness statement on 15 March 2019 along lines which were agreed
Counsel,2?° and on 18 March Gideon Cohen provided a schedule of findings in
the Common Issues Judgment to be fed into the draft statement.92° Amy Prime
sent the draft statement to David Cavender QC, Gideon Cohen and Stephanie
Wood on 19 March 2019 so that they could continue work on it.92” Gideon Cohen
responded indicating that, following a discussion with Lord Grabiner QC, the
Counsel team were considering not putting in a witness statement at all and
instead submitting only a (fuller) Application Notice.°?° WBD indicated that we
would be guided by the Counsel team’s views on whether the application should
be accompanied by a witness statement;9° in the end, Counsel advised that a
statement should be put in, albeit only a short one which should be considerably
stripped back from the version WBD had prepared.%°
960. The Counsel team circulated near final versions of the application documents on
20 March 2019, around an hour before Jane MacLeod communicated the
Board's approval to issue the application.*'! The witness statement which the
£25 WBON0000664.
826 WBON0001503; WBONO001504.
°27 POL00364150; POL00364151.
28 WBON0001512. The Counsel team were working on the draft Application Notice and draft
Order in parallel, see for example: WBON0000674; and WBON0001512.
828 WBON0000680.
930 WBONO001514.
°31 WBONO0001516. The witness statement circulated by the Counsel team is: WBON0001519.
Page 527 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Counsel team had prepared was finalised that evening,°°? and was sent to Jane
MacLeod along with the other application documents for consideration.°°>
961. The application was readied for issue and sent to the Court and Freeths, by which
time the HIT was already in session for its eighth day (with Torstein Godeseth
giving evidence). Anticipating that this would be the case, I recommended to Tony
Robinson QC not to proactively raise the subject of the Recusal Application
during the hearing. I felt it was preferable for his involvement in the Recusal
Application to be kept to a minimum, particularly given the possibility that the HIT
might proceed (and indeed I recall him having minimal input). Further, I thought
it would be more appropriate, and less awkward for Mr Justice Fraser, if he were
to find out about the application in his chambers rather than in open court. The
key consideration, in terms of the timing of the application, was that it should be
issued as soon as it was ready — both so that there could be no suggestion that
POL had not acted promptly, and so that it could be accommodated if possible
during the HIT’s non-sitting week the following week.
962. Mr Justice Fraser became aware of the Recusal Application during the lunchtime
adjournment and, once the cross-examination of Torstein Godeseth had finished
that afternoon, he listed a hearing for 3 April 2019 and directed (amongst other
things) that POL were to file a further witness statement by noon on Tuesday 26
March 2019: (i) giving further details of the specific findings in the Common
Issues Judgment on which POL relied as having been impermissible; (ii)
identifying the examples of the “critical invective” to which paragraph 25 of my
932 WBONO0001520.
33 POL00023769: see further Jane MacLeod’s initial response WBON0000685), Tom Beezer's
reply thereto (WBON0000686), and Jane MacLeod’s further reply (WBON0000687).
934 WBON0000684.
Page 528 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
fourteenth witness statement referred; and (iii) identifying the examples of the
criticisms of POL’s witnesses to which paragraph 25 of my statement referred.°°>
963. A first draft of that statement was prepared by Counsel and circulated on 24
March 2019,9°° and was added to by Amy Prime and Owain Draper.°°” My email
records indicate that on 25 March 2019, I had a call with Gideon Cohen (though
I cannot independently remember this call) in which I expressed the view that we
should adopt the approach of “includfing] anything [from the Common Issues
Judgment] that we might rely on, unless plainly not relevant to the application.
The objective being to avoid any challenge on appeal that a point in the Judgment
was not put to Fraser on the application”.**® Gideon Cohen responded to the draft
which WBD had prepared in line with this approach, highlighting some passages
that he felt it was better not to include as they were borderline. Lord Grabiner
QC endorsed Gideon Cohen’s approach in the following terms:
“If Gideon thinks they're problematic please leave them out. We've got a
lot of good material and I don’t want it undermined by taking points that
don't really work’.94°
964. I accepted that advice! and the draft witness statement was then sent to the
POL legal team on the morning of 26 March 2019.%? Amy Prime’s covering email
set out that the approach had been to only focus on material from the Common
Issues Judgment,%3 and “to include in the statement as much as possible from
935 WBONO000688.
896 POL00364171; POL00364172.
887 POL00364173; WBON0000689.
838 POL00364175.
838 POL00364177.
£40 WWBONO000692.
°41 WBONO0000693. Further amendments to the statement were suggestions by Counsel and
incorporated into the statement: POL00364183; WBON0000694; WBON0001522.
42 POL00023950.
°43 As to this, POL were keen for excerpts from the HIT transcripts to date which were arguably
indicative of bias on Mr Justice Fraser's part to be included: WBON0001523. Counsel advised
against this approach on the basis that “/the] risks to us of this application sprawling into a
Page 529 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
the Judgment so as no points are missed, which has been weighed against the
risk of including such a large volume of material that we are accused of being
unhelpful’. The statement was approved by POL and filed and served that in
advance of the hearing of the application on 3 April 2019.94
Mr Justice Fraser's Recusal Judgment (Q114)
965. By Q114, I am asked to consider Mr Justice Fraser's judgment dated 9 April 2019
(the “Recusal Judgment’). I do not recall specially when I considered this
judgment, but I expect I would have considered it on the day I received it.
966. I was not surprised by the grounds upon which Mr Justice Fraser rejected the
application as they were all points which Counsel had anticipated when advising
on the merits of the application, and I understood from Counsel that it was to be
expected that Mr Justice Fraser would refuse the application in the first instance.
I address the main points Mr Justice Fraser made further below in the context of
responding to Q118, since Lord Justice Coulson relied upon essentially similar
points in rejecting POL’s application for permission to appeal the Recusal
Judgment.
967. I do however wish to briefly remark on [122]-[123] of the Recusal Judgment. Here
Mr Justice Fraser criticised (i) the fact that my thirteenth witness statement, which
general complaint about his behaviour throughout the litigation are significant. We will have to
work pretty hard not to sound like a disappointed litigant: WBON0001531.
°44 POL00023239. For completeness, I add that later on 26 March 2023, Freeths wrote a letter to
WBD complaining that my fifteenth witness statement did not exhaustively list all of the
passages of the Common Issues Judgment which POL considered supported its case on
recusal: POL00269583. They expressed the view that this was in breach of the directions Mr
Justice Fraser had given. Counsel advised that we resist this suggestion, as the statement had
struck an appropriate balance and it was open to the Claimants to object at the hearing if we
relied on passages that were not referenced in the statement: WBON0000695; WBONO000696;
POL00269584. See further: WBON0001696; WBON0000697; WBON0000698; and my views
at WBONO0000699. The Claimants thereafter requested a short hearing to debate this point,
which took place on 27 March 2019: WBON0000700. No further Order was made by Mr Justice
Fraser at this hearing.
Page 530 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
was filed and served the same day as the Recusal Application, did not mention
that the application was being made; (ii) the fact that the application was not
foreshadowed in open court on 21 March 2019; and (iii) the fact that it was made
some two weeks after POL had received the draft Common Issues Judgment. As
to (i), my thirteenth witness statement was prepared separately from the Recusal
Application and related to an entirely different topic (see above, §693) and so I
would not have considered it necessary or appropriate to cross-refer to the
application in that statement; and as to (ii), I have explained the reasons why
application was not mentioned in open court above at §961. In both instances, I
consider that the approach taken was reasonable; it was certainly not done to
‘keep the application up one’s sleeve’. As to (iii), as I have explained above and
as we were advised by Lord Neuberger, there were restrictions on what could be
done whilst the Common Issues Judgment was under embargo (which it was
until 15 March 2019). Thereafter, it took less than a week for POL to obtain
advice, for the Board to take a decision, and for the Recusal Application to be
finalised and issued. The application was made the working day after POL’s
Board gave its approval. In the circumstances, and overall, I consider that it was
made and brought to the Court’s attention as soon as it reasonably could have
been.
(v) Decision-making after the Recusal Judgment (Q115 to Q117)
968. Q115 to Q117 relate to events after the Recusal Judgment was handed down.
Email to Jane MacLeod of 14 April 2019 (POL00023208; Q115,
969. By Q115, I am asked to consider POL00023208, which is an email I sent to Jane
MacLeod on 14 April 2019, collating various emails from Lord Neuberger and
Page 531 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Lord Grabiner QC concerning the approach they thought POL should take to
seeking permission to appeal the Common Issues Judgment and Recusal
Judgment.
970. By this time, an application for permission to appeal the Recusal Judgment had
been lodged, and Counsel's view was that this application should be dealt with
alongside any application for permission to appeal the Common Issues
Judgment. Since POL had not yet taken a decision on whether to appeal the
Common Issues Judgment, this raised a question as to whether permission
should be sought directly from the Court of Appeal, and whether the Listing Office
should be notified so that the application in respect of the Recusal Judgment was
not progressed in the meantime. Earlier on 14 April 2019, I had asked the
Counsel team to prepare draft letters to the Listing Office for POL to consider,
and sought their availability for a call the following day to advise Jane MacLeod
on their suggested approach.%5 I then sent Jane MacLeod the email in
POL00023208, in which I collated some of their emails on these matters to date.
Following this I received further emails from Lord Neuberger and Lord Grabiner
QC laying out their recommendations. I compiled these into a further email to
Jane MacLeod which I sent on the morning of 15 April 2019. Since their advice
was clear and direct, I indicated to her that I no longer thought a call was
necessary to obtain their advice on how to proceed.™® I subsequently confirmed
to Counsel that a call was not required.24”
Email to Jane MacLeod of 17 April 2019 (POL00006513; Q116)
£45 WWBONO000706.
£46 WWBONO001572.
°47 WBONO0000707.
Page 532 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
971. Q116 asks me to consider an email I sent to Jane MacLeod on 17 April 2019
(POL00006513). Specifically, I am asked by Q116.1 to explain the first bullet
point in that email, in which I informed Jane MacLeod that Lord Neuberger and
Lord Grabiner QC had cautioned against conceding implied terms in any appeal
against the Common Issues Judgment as “they fear[ed] that making concessions
is a slippery slope, and even making a few concessions indicates to the Court of
Appeal that the SPM contracts are incomplete and open to further implied terms.
They therefore advise that the implied terms that DCQC suggested be conceded,
should only be offered to the Court as incidents of the ‘necessary cooperation
term”.
972. The background to this was that, following the hand-down of the Common Issues
Judgment and whilst work on the Recusal Application was ongoing, David
Cavender QC was working on draft grounds of appeal against the Common
Issues Judgment. In the course of that work, David Cavender QC and I discussed
the possibility of POL accepting certain of the implied terms contended for by the
Claimants (or modified versions of them), on the basis that POL’s case in respect
of some of the terms was stronger than others; the Court of Appeal might be less
likely to grant permission if POL were seen to be challenging every aspect of the
Common Issues Judgment; and conceding certain terms would help to help to
send the message that POL was being careful and reasonable in its approach.
David Cavender QC, for example, was in favour of conceding a version of implied
term (t) (to the effect that POL had a duty to take reasonable care in carrying out
functions which could affect SPMs’ branch accounts);*48 and I suggested
conceding implied terms (a) and (b) (requiring POL to provide adequate training
48 WBON0000158.
Page 533 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
and support and to ensure that the Horizon system was reasonably fit for
purpose).°4° Making concessions where possible was also something that POL
were keen to explore at the time.9°
973. On 10 April 2019, David Cavender QC sent me draft grounds of appeal which
proposed conceding the above terms, as well as a version of term (c) (requiring
POL to properly and accurately effect transactions using Horizon and to keep
records of such transaction), in each case on the basis that these terms were
facets of “necessary cooperation”.°*' This approach was also reflected in a draft
paper we prepared and sent to POL the same day for discussion at a conference
with Alisdair Cameron and others scheduled for 11 April 2019 (see §§3.2 to
3.3);95* and in a draft schedule prepared following that conference, identifying
which points from the Common Issues Judgment POL should and should not
appeal.®° It was an updated version of this schedule that I sent to Jane MacLeod
as an attachment to my email of 17 April 2019 (POL00006513).°*
974. Lord Neuberger had, at this time, been instructed to consider the scope of the
appeal against the Common Issues Judgment and to that end was sent David
Cavender QC’s draft grounds of appeal on 10 April 2019.9° My email records
indicate that Lord Neuberger thereafter discussed the scope of the proposed
appeal against the Common Issues Judgment with David Cavender QC, as did
°49 WBON0000200.
950 See for example: WBON0000323and POL00023941.
951 POL00270456; POL00270457(§§30, 32, and 37). On reflection, I came to the view that this
concession may need to be modified or dropped, see: WBON0001575.
°52 POL00023028; POL00270458.
953 WWBON0001576; POL00270870. This draft also suggested conceded implied term (r), namely
that POL must not exercise its contractual powers in an arbitrary or capricious manner.
54 POL00270936.
855 WBON0001541.
Page 534 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Lord Grabiner QC.9* On 16 April 2019 (in response to some reflections I had
made on the proposed acceptance of implied term (c)), David Cavender QC said
that the “approach urged by DN and AG is not to actually concede any of these
implied terms — whether as narrowed by me or at all. And to seek to hold the line
at Necessary Co-Operation and Stirling v. Maitland and then make warm noises
about the possibility of some of these "necessary" implied terms (if sufficiently
defined /narrowed as i have suggested) coming in under the umbrella of those
Agreed Implied terms- in the skeleton argument’ .°°” This is what appears to have
informed the first bullet point in my email in POL00006513, i.e. noting that Lord
Neuberger and Lord Grabiner QC had warned against conceding any implied
terms (and advising that if any were to be conceded, this should only be on a
‘necessary cooperation’ basis as David Cavender QC and I had suggested).
975. I note that Lord Neuberger and Lord Grabiner QC remained of this view; when
advising on POL’s updated draft grounds of appeal (which no longer expressly
conceded any implied terms), Lord Grabiner QC commented:
“David and I think the draft covers the points. We do not think that
concessions should be made, eg on the implied terms, because, as
previously advised, we think the co-operation and Stirling v Maitland
implications are effective and would be readily implied in this case
without the need to manufacture further terms which is what the Judge
in his wisdom has done” .%*
976. With reference to Q116.2, based on searches of my email records and calendar
appointments I do not believe I had any conferences with Lord Neuberger or Lord
Grabiner QC about the Recusal Application other than: (i) a meeting I attended
956 See for example WBON0001541.
987 WBON0001575.
858 WBON0000291; cf. POL00284926.
Page 535 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
with Lord Grabiner QC on 2 April 2019 in advance of the Recusal Application
hearing the following day; and (ii) the Subcommittee meeting on 24 April 2019
(discussed below). The meeting on 23 April 2019 to which POL00006513 refers
was to discuss the prospective appeal against the Common Issues Judgment. I
had a number of meetings and calls with David Cavender QC during this period
(including some with POL client contacts), however I cannot see from those
emails or appointments that any of them were for the specific purpose of
discussing the Recusal Application as opposed to wider aspects of the case in
which David Cavender QC was involved (for example, the proposed appeal
against the Common Issues Judgment).
Subcommittee meeting on 24 April 2019 (Q117)
977. By Q117 I am referred to the minutes of the POL Board Subcommittee meeting
on 24 April 2019 (POL00006755). I do not believe that I have seen these minutes
before and this appears to be confirmed by their absence from my firm’s file. I
recall that the purpose of this meeting was to discuss the tactical approach to the
appeals against the Recusal Judgment and the Common Issues Judgment,
specifically whether they should be joined and run together or should proceed
separately. I recall that HSF, who by this time had been brought in to advise the
board, suggested keeping the appeals separate, whereas the Counsel team’s
view was that it was more coherent and likely to be tactically advantageous for
them to be dealt with together. I do not recall any other differences of opinion,
although my recollection of this meeting is impressionistic rather than detailed
and I do not recall speaking at it (and for these same reasons, I have no reason
to doubt the accuracy of the minutes). To the best of my recollection, there was
no disagreement as to the fact that (i) the application for permission to appeal
Page 536 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
the Recusal Judgment (which by this time had been lodged) should be
maintained; and (ii) that an application for permission to appeal the Common
Issues Judgment should be made. In respect of both appeals, my recollection is
that POL were primarily merits-driven and their key concern was therefore
whether the prospects of success were good. As such, POL were content to
accept Counsel’s clear advice that both appeals were justified on the merits
(though I recall the Subcommittee asking probing questions about the recusal
appeal in particular). As to the Common Issues appeal, as I have explained
above, the main consideration was not whether an appeal should be brought, but
whether it should be linked to the recusal appeal, how wide it should go and
whether any concessions should be made as to the terms to be implied into
POL's contracts with SPMs.%°
(v) Refusal of permission to appeal the Recusal Judgment (Q118)
978. By Q118, I am asked to consider POL00023207, which is Lord Justice Coulson’s
decision dated 9 May 2019 (the "PTA Decision") refusing POL’s application for
permission to appeal the Recusal Judgment. I am asked to set out whether, on
reflection, I accept a number of criticisms that Lord Justice Coulson made therein.
979. First, I address Lord Justice Coulson's criticism that POL’s application
misrepresented the facts in dispute during the CIT (§§21-23, Q118.1). I do not
accept this suggestion; in particular, I do not agree that the Recusal Application
959 I ater, towards the end of May 2019 after Lord Justice Coulson had refused permission to appeal
the Recusal Judgment, Helen Davies QC was instructed to lead on the Common Issues appeal.
I recall that she advised that the scope of the proposed appeal should be narrowed so as to
remove the procedural unfairness grounds. By this time however, HSF were primarily advising
POL on the appeal strategy.
Page 537 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
was based on factual findings made by Mr Justice Fraser which POL itself had
put in issue:
979.1.There was in my view a very significant difference between the Claimants’
case and POL’s case in relation to ‘post-contractual’ matters, namely that
POL considered those matters inadmissible and that post-contractual
factual findings were not needed to resolve the Common Issues (save in
a couple of very narrow respects addressed for example at fn. 758);
whereas the Claimants actively invited findings on these matters. This was
consistent with the witness statements that POL served where it
repeatedly explained that its evidence was on matters it expected an SPM
to know at the time of contracting. For example, Angela Van Den Bogerd's
evidence on training was expressed to be limited to the "level of
information to be known by an applicant from their own enquiries when
applying to be a Subpostmaster or to be communicated to an applicant
during the appointment process".
979.2.In relation to the two examples given in the PTA Decision (§21):
(i) The PTA Decision refers to Nick Beal giving "a good deal of evidence
about the NFS/P]". However, his witness statement only included
seven paragraphs about the NFSP which explained who the NFSP.
were and what they did; being the type of information that a person
would likely know or could have found out when applying to be an
SPM.
(ii) I accept that findings were sought on Mr Abdulla's use of the NBSC
helpline in POL's written closing submissions, but that was only in
Page 538 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
relation to credit and was later withdrawn — I address this further below
at §981.4 and §982. In the end, no part of POL's case on the Common
Issues relied on factual findings being made about how SPMs used
the helpline in practice.
979.3.Looking back now at the witness statements that POL served, I do think
the evidence of Helen Dickinson on the risks of fraud in the branch network
could have been better explained to make clear that it was intended to
describe the general risk of fraud in any retail-type business faces, and
thus formed part of the objective background facts known to an SPM at
the time of contracting. I also think, on reflection, that the last five
paragraphs of Angela Van Den Bogerd's statement on the topic of
responsibility for shortfalls went over the line of admissibility.
980. Second, I address Lord Justice Coulson's criticism that POL itself put issues into
dispute through its cross-examination of the Lead Claimants (PTA Decision, §25,
Q118.1). As to this, there is in my mind an important difference between POL
leading evidence through its own witness statements and cross-examining the
SPM's evidence. All of the Lead Claimants gave extensive evidence in their
witness statements on events that took place after they became SPMs. I do not
believe that POL should have been criticised for cross-examining on post-
contractual facts that the Claimants themselves put into evidence through their
statements. POL was entitled to challenge the accuracy of the Claimants’
evidence whilst also maintaining that it was inadmissible and not seeking findings
on those matters. Similarly, the Claimants’ Counsel elected to cross-examine
POL's witnesses on post-contractual matters, but it would be wrong to
characterise that as POL putting those points into dispute.
Page 539 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
981. Third, I address Lord Justice Coulson's point that it was appropriate for Mr
Justice Fraser to make findings of fact in respect of credit arising from the Post
Office’s cross examination (§§24-26, Q118.2). I have explained the approach
POL took to the cross-examination of the Lead Claimants in the CIT at §§765-
678 above in response to the Inquiry’s Q90.4. In POL's written closing
submissions, the key point on which POL said that the Claimants’ evidence might
be relevant was the question whether they had received a copy of their SPM
contract (paragraph 560). POL's primary position was that much of even this
evidence was irrelevant because, other than Ms Stubbs, each Lead Claimant
had signed a document accepting their contract terms (paragraph 570). Its
alternative position was that the Claimants had, in fact, received their contract
terms and in that respect made a few limited submissions as to credit:
981.1 Of the six witnesses called by the Claimants, no findings as to credit
were sought in respect of three: Ms Stubbs,°°° Ms Stockdale,°°' and Ms
Dar.962
981.2 In relation Mr Bates, POL's submissions as to credit went to the timing
of when he had received his SPM contract. Those submissions were based
on pre-contractual events around the time of his appointment, save for one
letter that Mr Bates sent after he was appointed as an SPM which referred
to his contract as being "very wordy" indicating that he did have a copy of
the contract.
960 POL’s written closing submissions, paragraph 578.
961 Ibid, paragraph 597.
962 Ibid, paragraph 598.
963 Ibid, paragraph 577(b).
Page 540 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
981.3 In relation to Mr Sabir, there was a single sentence that made a minor
comment on credit — this entirely turned on pre-contractual events.°%*
981.4 There was a single paragraph submission as to credit in respect of Mr
Abdulla, which in part referenced post-contractual events.°®> However, POL
subsequently refined its position in respect of Mr Abdulla. Following the
filing of its written closing submissions, Mr Justice Fraser invited a
submission from POL about how it should address findings as to credit.
This was filed on 17 December 2018.9
"Post Office’s position is that in making those findings, and taking
that view on credibility, the Court should:
(a) Take account of evidence given by witnesses on matters
within the scope of the Common Issues trial. So, for example, the
Court's findings on whether Mr Bates received a copy of the
SPMC will presumably take into account the evidence he gave on
that issue, and on associated issues raised in cross-examination
(for example, whether he is careful generally, whether he had a
copy of the SPMC when writing to Post Office in August 1999,
and so on).
(b) Take account of evidence on matters which go to the
witnesses’ credibility, but do not risk trespassing on any future
trial, because they do not go to issues of breach or causation. For
example, Mr Abdulla’s evidence on whether Christine Adams and
Christine Stephens were the same person can be taken into
account in assessing his credibility.
(c) Not take account of evidence which, while it may go to the
witness's credibility, risks trespassing on a future trial or trials. For
example, the Court should not make any findings on whether Mr
Abdulla falsely accounted, even though such matters might be
relevant to his credibility. Nor (staying with this example) should
the Court base any findings on Mr Abdulla’s credibility which are
964 Ibid, paragraph 589: "His evidence that he thought the ‘standard’ term contract in fact contained
his obligations to do certain things at the branch in his first 6 months as SPM was not credible".
°65 Ibid, paragraph 592.
886 WBON0001717; POL00259980.
Page 541 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
necessary to decide the Common Issues on his evidence as to
the allegations of false accounting made against him."
982. I therefore do not believe that POL's submissions on credit invited Mr Justice
Fraser to make findings of fact on post-contractual matters (save for the limited
point set out at (b) above in respect of Mr Abdulla, which POL maintained for the
reasons set out therein). I can understand why POL’s approach may have
appeared confusing, because POL did cross-examine on post-contractual
matters and then, in respect of Mr Abdulla, did initially seek a very limited number
of post-contractual factual findings that went to his credit.
983. Fourth, I deal with Lord Justice Coulson’s criticism that it was appropriate for Mr
Justice Fraser to make the findings of fact that he did, including in respect of
post-contractual matters (§§27-28, Q118.3). As I have explained above (e.g.
§§699, §712.3, §714), and as was reflected throughout the advice WBD and
Counsel provided to POL before, during and after the CIT, there were a small
number of areas where it was accepted that limited evidence on post-contractual
points was permissible. The basis of the Recusal Application was that Mr Justice
Fraser had made factual findings on post-contractual matters that went far
beyond that permitted even allowing for those areas.
984. Fifth, I address Lord Justice Coulson’s point that the fact a judge may make
unnecessary findings of fact would not give rise to apparent bias, unless it
amounted to actual or apparent pre-judgment (PTA Decision, §30, Q118.4). I
agree with this, and indeed this point had been carefully considered in the advice
on recusal that was provided to POL. For example, it was dealt with in Lord
Neuberger’s note of advice dated 14 March 2019, at paragraph 18. In the instant
Page 542 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
case, the basis of the Recusal Application was that Mr Justice Fraser’s
unnecessary findings of fact had amounted to pre-judgment.
985. Sixth, I address the charge that POL was “wholly unjustified” in criticising Mr
Justice Fraser’s repeated declaration that he was not making findings on breach
etc. (PTA Decision, §§31-36, Q118.5). I accept that there is some nuance here.
In the Common Issues Judgment, Mr Justice Fraser did explain that he was not
making any findings as to breach, causation, or loss; and I recognise that in later
trials he would not have been strictly bound to the observations he did make.
However, overall I felt (with the benefit of Counsel’s advice) that the number and
nature of the observations he made, combined with the fact that they frequently
went to matters of breach and liability, indicated that he would find it difficult to
approach some of those issues with an open mind later on. To be clear, to my
mind this did not amount to an accusation of bad faith on Mr Justice Fraser’s part
(cf. POL00023207 at §34).
986. Seventh, I consider Lord Justice Coulson’s point that Mr Justice Fraser was
entitled to form a critical view of witnesses (PTA Decision, §§41-45, Q118.6). I
accept that a judge is entitled to form a critical view of witnesses, including (for
the avoidance of doubt) where that witness is due to appear before them in a
later trial. However, I did feel (and Counsel agreed) that the language Mr Justice
Fraser adopted in relation to POL and its witnesses was excessively critical and
that the phrase “critical invective”, whilst strong, fairly reflected this. That said,
the reference to Mr Justice Fraser’s language and “critical invective” was the part
of the recusal application I was least enthusiastic about. I did not consider that it
added any substance (and as I recall, it was covered in a single paragraph of my
Page 543 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
fourteenth witness statement in support of the Recusal Application).°°” To the
extent that there was a point to be made about Mr Justice Fraser’s treatment of
POL’s witnesses, I felt that the strongest aspect of the argument was that he had
criticised them for not covering certain topics in their statements, when they had
excluded these matters on the advice of POL’s lawyers.9°°
987. Eighth, I address Lord Justice Coulson’s criticism that the Recusal Application
had ‘no substance’ (PTA Decision, §50; Q118.). I disagree with this point. As I
have set out above at §946.5, the Recusal Application was based on the advice
of Lord Neuberger, Lord Grabiner QC, and David Cavender QC, each of whom
advised that there were (at the very least) reasonable grounds for making the
application.
988. Overall, whilst I acknowledge that Lord Justice Coulson considered that the
Recusal Application was not well-founded, in my view it was not unreasonable
for POL to have made the application or indeed to have sought leave to appeal
its refusal. POL had the benefit of strong advice from both Lord Neuberger and
Lord Grabiner QC to the effect that (i) the application was well-founded, and (ii)
an appeal against its refusal should be brought (and by this time both Lord
Neuberger and Lord Grabiner QC had been acquainted with the case for around
a month).°® This view was also held by David Cavender QC, who was trial
Counsel and had a deep knowledge of the underlying factual basis for the
application, as well as our supporting cast of junior Counsel. Indeed, even after
Lord Justice Coulson’s PTA Decision both Lord Neuberger and Lord Grabiner
967 POL00269105, paragraph 25.
968 Ibid, paragraph 29.
69 On POL's prospects of appealing the Recusal Judgment, see for example WBON0000169 and
WBONO000172.
Page 544 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
remained of the view that Mr Justice Fraser had erred in his Common Issues
Judgment, and had done so in such a way as to indicate that he ought to have
been recused. Having reflected on the Order, Lord Neuberger said:9”°
“The clients will naturally feel both disheartened by the judgment and
bemused by the fact that the view taken by an appeal court judge is
entirely inconsistent with that of their legal advisers.
As to being disheartened, the main appeal (on interpretation) is
unaffected by the Coulson LJ judgment, at least in any direct sense. In
case the Coulson LJ judgment is thought to be relevant to the main
appeal, my experience over 45 years shows that successive setbacks in
litigation come in two categories: (i) those which should make you realise
that you are on the wrong track, and (ii) those which should stiffen your
resolve. It is of course normally easy when it is all over to identify which
category you were in, but harder to do this when one is in the middle of
the litigation. Having said that, on the main interpretation issues, I remain
firmly of the view that we are a category (ii) case. The issues actually
decided by Fraser J involve applying what I regard as well-established
principles of law, and in that connection I think he has gone seriously
wrong. The reasons for my view are all to be found in the recently
prepared grounds of appeal and skeleton argument.
As to being bemused, when it comes to the recusal appeal we are in a
more nuanced area of judgement, and there is, I acknowledge, at least
in principle, a greater risk of this being a category (i) case. That was my
main reason for leaving it to stew overnight. Having done that, I remain
of the opinion that Fraser J should have been recused, despite the fact
that Coulson LJ and Fraser J disagree: neither their reasons nor their
identity has caused me to change my view.”
989. At the same time, and notwithstanding the strong advice received from Lord
Neuberger, Lord Grabiner QC and David Cavender QC, WBD and POL were
alive to the sensitivities attendant on making the Recusal Application (as were
Counsel). Conscious effort was therefore made to bring the application to the
Court's and Claimants’ attention as soon as practicable after hand-down of the
Common Issues Judgment, and close attention was paid to matters which could
970 WBONO000148.
Page 545 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
give rise to any accusations of heavy-handedness, such as the tone of our
correspondence with Freeths and the tactical approach to the appeals.
R. SUBSEQUENT EVENTS (INCL. Q119 and Q120)
990. This section relates to the period following the end of the HIT. Specifically, it
addresses: (i) the extent of my/WBD’s role in respect of criminal appeals at this
time (Q119), (ii) the discovery and disclosure in Autumn 2019 of additional KELs
which had not been disclosed to the Claimants prior to the HIT, and (iii) my role
following the hand-down of the Horizon Issues Judgment (Q120).
(i) Criminal appeals (Q119)
991. By Q119 I am asked to what extent WBD were instructed to advise POL in
relation to criminal appeals, with reference to POL00022933 (which is an email
chain in March 2019 in which Amy Prime provided instructions to Brian Altman
QC to consider the Common Issues Judgment and advise on whether it
undermined the safety of historic criminal convictions).
992. WBD were not instructed to advise POL on these matters. Our involvement was
limited to seeking advice from Brian Altman QC as set out in POL00022933, in
keeping with our historic role as his instructing solicitors. His advice was received
on 14 April 2019 and provided to POL the same day.°”1 From mid-2019 Brian
Altman QC’s instruction was directed by HSF, and I was aware that when the
Horizon Issues Judgment was delivered he was (as with the Common Issues
Judgment) asked to consider it and to provide his views on its impact on historic
871 WBON0001697; POL00273923; POL00023115.
Page 546 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
criminal convictions.9’2 In connection with this, WBD collated information
requested by Brian Altman QC where required to do so.9”8
(ii) Late disclosure of back-versions of KELs
993. Although not arising directly out of the Inquiry’s questions, there is one matter
which I have touched upon above (at §522 and §551.3) which I wish to address
in a little more detail. This concerns WBD’s discovery, in Autumn 2019, of the fact
that Fujitsu held copies of historic versions of KELs which had not been disclosed
to the Claimants (although the most recent versions of the relevant KELs had).
994. In summary, this came about as follows. In September 2019, POL was putting in
place a new process for monitoring issues with Horizon and as part of that
process POL and WBD became aware of a number of KELs and Peaks which
were potentially relevant to the Horizon Issues, but which had not been disclosed
as they were recently created. These documents were disclosed to Freeths
under cover of a letter dated 25 September 2019.94
995. Freeths’ response included a question about whether KELs which we had
previously disclosed in January 2019 had been updated, such that new, more
recent versions of these KELs now needed to be disclosed.°”> In response, we
referred Freeths to our description of the KEL in the EDQ, which had stated:
“[t]he KEL only contains the current database entries and is constantly
updated and so the current version will not necessarily reflect the version
that was in place at the relevant time. The previous entries / versions of
the current entries are no longer available”.
972 WBON0001656; WBON0001660; WBON0000721.
873 WBON0000723; WBONO000724.
°74 WBONO0001653.
975 POL00285257.
Page 547 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
996. That statement was based on information provided by Fujitsu and had been
approved by them prior to the EDQ being filed in December 2017. However, since
that time, we had been given reason to be less confident about the reliability of
Fujitsu’s instructions. With this in mind, Amy Prime reviewed our draft response
to Freeths and commented: “This needs double-checking with [Fujitsuy’ .9”°
997. Fujitsu were asked to confirm the wording above on 30 September 2019.97” On
1 October 2019, Matthew Lenton of Fujitsu responded as follows:
“This part: “[t]he KEL only contains the current database entries’— I’m
not completely clear what that is intended to mean, but it may be clarified
by the following: This is correct: “is constantly updated and so the current
version will not necessarily reflect the version that was in place at the
relevant time” The second sentence is not correct: “The previous entries
/ versions of the current entries are no longer available". You may recall
that there are three status categories of KEL: current, deprecated and
deleted. For those that are current or deprecated, they have been
updated in such a way that previous content is not permanently
overwritten, but instead a new version is created, with the previous
versions being retained and accessible. For those_that_have_ been
deleted, only the last version at the point of deletion has been retained”
(emphasis added).278
998. This was a surprise to me and my team. On the face of it, the statement given in
the EDQ was incorrect, and we had also not provided disclosure of historic
versions of some KELs which had been relied upon by the parties and their
experts at the HIT. Amy Prime spoke with Matthew Lenton on the phone that day
and took a record of his explanation:
“When Fujitsu revise the contents of an existing KEL they would not
overwrite the KEL but take a copy of the KEL, make the changes and
Save as a new document. The previous versions of the KELs would be
kept for version control and sit underneath ... The previous versions are
876 WBON0000324.
°77 WBONO0000325.
878 POL00043025.
Page 548 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
held in the same database as the latest version of the KELs. They are
not actively archived off to a different location. When providing us with
the documents for disclosure, Fujitsu just extracted the latest version of
each KEL. Without doing some further investigations, Fujitsu do not
know whether all sequential KEL versions are there.
o Current and deprecated KELS should have previous versions but
would need to double-check
o Deleted KELs were flattened, meaning that the previous versions
which were held under the latest version were “knocked out”.°”°
999. On 2 October 2019, Amy asked Matthew Lenton to give an estimate as to how
long it would take to extract historic versions of KELs and on 4 October 2019, I
followed up with asking for an estimated timeframe as a matter of urgency. We
also took steps to satisfy ourselves that other documents produced by WBD had
not contained the same inaccurate statement as the EDQ. We took advice from
Counsel who confirmed our view that the inaccuracy needed to be drawn to
Freeths’ and the Court's attention promptly.
1000. We therefore wrote to Freeths on 3 October 2019 in the following terms:
“In respect of your query regarding the fact that intervening changes to
KELs are not captured on the face of the KELs disclosed, we have made
further enquires with Fujitsu to confirm our understanding that previous
versions of the KELs are no longer held. We regret to say that these
enquiries have revealed that our understanding was wrong. As you will
be aware, POL’s EDQ stated that “[t]he KEL only contains the current
database entries and is constantly updated and so the current version
will not necessarily reflect the version that was in place at the relevant
time. The previous entries / versions of the current entries are no longer
available”. This statement was based on info provided by Fujitsu. In
response to our recent enquiries, however, we were informed by Fujitsu
on 30 September 2019 that this is incorrect and that previous versions
of KELs are available. This takes POL greatly by surprise. It relied on the
information provided by Fujitsu at the time of filing its EDQ that such
documents were not available. It is extremely sorry that this information
878 WBONO0000137.
880 POL00043025.
Page 549 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
was incorrect. Regardless of whether these document may or may not
be "adverse", POL has taken immediate steps towards arranging for the
previous versions of these KELs to be extracted and has instructed
Fujitsu to begin producing a script to extract the documents into a
readable HTML format. Unless you disagree, once this script has been
produced we will ask Fujitsu to extract all of these versions.
When this has been done, again unless you disagree, we propose to
take immediate steps to disclose previous versions of the KELs which
were referred to by either Dr Worden or Mr Coyne in their expert reports
or the joint statements or the bug table, or were for any other reason
included in the trial bundle. Disclosure of the previous versions of these
KELs (where there are previous versions) will be given as a matter of
urgency. We invite you to tell us whether the Claimants also wish to be
provided with disclosure of previous versions of the KELs which were
not referred to by the experts or included in the trial bundle. Given the
seriousness of this matter, we propose to notify the Managing Judge of
this development immediately and will send him a copy of this letter
under cover of an explanatory email”.°8"
1001. We informed the Court in similar terms on the same day,°® and started work on
what would have been my twentieth witness statement in the expectation that
the Court would order POL to file a witness statement to address the
misstatement in the EDQ. In the event, the Court did not seek any further
explanation of this issue and so this draft witness statement was not finalised
and served.
1002. Disclosure to Freeths of an additional 346 KELs (being prior versions of the
KELs in the HIT Bundle) was made under cover of a letter dated 24 October
2019. That letter also set out that we estimated that there were around 5,000
previous versions of KELs in total. We offered these further KELs to Freeths if
they wished, but they did not ask for them. I inferred from this that they did not
consider them useful — they were only earlier versions of KELs they already had
°81 POL00285691
982 POL00285690.
Page 550 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
— or Mr Coyne was already aware that back versions were available from when
he inspected the KEL database at Fujitsu's offices (where I understood the back
versions would have been visible on screen).
1003. My understanding at the time the EDQ was filed and served was that previous
versions of existing KELs were not available. The EDQ was prepared on that
basis and sent to Fujitsu for approval (who commented on our description of the
KEL database but did not object to the statement that “The previous entries /
versions of the current entries are no longer available”). As soon as we
discovered the error, we took steps to correct the position and to provide
disclosure to the Claimants of what seemed to us to be the most important
category of historic KELs. For completeness, in the course of reviewing my firm’s
records following the end of the group litigation I have identified an email from
Matthew Lenton dated 13 February 2019 in which I was in copy, which made
reference to “deprecated KELs”, identifying them as “superseded versions of live
or deactivated KELs” and stating that Fujitsu had “only provided the most recent
versions, so we have not separately provided these superseded versions” .°*
This email was in the context of an enquiry about a different category of KELs
(known as 'deleted' KELs) to which Mr Coyne had referred in Coyne 2, which Mr
Coyne said had not been disclosed. Fujitsu confirmed that the KELs to which Mr
Coyne referred had all been deleted and therefore were “no longer retrievable”;
and their reference to historic versions of live and ‘deprecated’ or ‘deactivated’
KELs was incidental to this, and I am confident that it was simply not picked up
by my team at the time (their focus being on dealing with Mr Coyne’s query about
°83 POL00285786.
884 WBON0001430.
Page 551 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
deleted). I am certain that I was not aware of the availability of back versions of
KELs, or the fact that they had not been extracted and provided by Fujitsu, prior
to this matter being drawn to my attention in early October 2019 (after which
steps were immediately taken to rectify the position).
(iii) Work since the Horizon Issues Judgment was handed down (Q120)
1004. Following the hand-down of the Horizon Issues Judgment on 16 December
2019, my involvement largely came to an end. The GLO had by this time settled
following a nine-day mediation, and HSF were the lead lawyers for Horizon-
related matters (which had been the case from around mid 2019).
1005. During the period from hand-down of the Horizon Issues Judgment until 26
February 2020, the further work WBD performed for POL “in relation to the
Horizon IT system” (which I take to include POL’s related accounting and
business practices) largely consisted of:
79.1 Supporting HSF, as requested, in its post-settlement work on the group
litigation by providing information and offering my views where required
(for example, in relation to costs issues, the discharge of the GLO, and
the implementation of the settlement agreement e.g. by assisting in having
Claimants’ bankruptcies annulled).
79.2 Providing occasional input on POL’s ongoing programme of work to
modify its business practices in order to comply with the Common Issues
Judgment.
79.3 Providing information and documentation to HSF and Peters & Peters
(POL's new criminal lawyers) based on WBD’s historic involvement in the
Page 552 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
Horizon-related matters. A number of the requests to which we responded
during this period were in connection with dispute resolution processes
and other remediation activities that (as I understand it) fall outside the
terms of POL’s waiver of privilege.
79.4 Other work falling outside the terms of POL’s privilege waiver.
1006. Overall I would estimate that the proportion of my time spent on this work was
less than 10%.
S$. CONCLUSION (INCL. Q121 and Q122)
1007. I conclude my statement with some observations by way of overview and
conclusion, including with reference to the Inquiry’s Q121 which asks, “fijn
hindsight, is there anything that you would have done differently in respect of the
matters raised in your statement’?
1008. Throughout this statement I have endeavoured to acknowledge where mistakes
were made by me or my firm and where misunderstandings arose, and to explain
how these came about and what was done to rectify them. For example, the
disclosure of E&Y’s audit reports (at §§672-694), the evidence around counter
injections (at §§841 ff), the late disclosure of Peak PC0273234 (at §§631-647),
etc.). I believe that in each of these instances, these were genuine mistakes and
that I and my team acted appropriately and in good faith. In my view it is not
surprising that there were some mistakes and misunderstandings given the
scale, complexity and the pace at which we were working.
1009. In this regard, and as a general point, I consider that it would have been helpful
to have had a second matter Partner instructed for POL, at least during 2018
Page 553 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
when we were preparing simultaneously for both the CIT and the HIT. I do not
think this was necessary for the duration of the group litigation; on the whole, the
size and structure of the WBD team worked well, albeit that we were busy. It can
be difficult to decide on the structure of a team for a large piece of litigation, and
we considered the matter with care and kept it under review. However, the
structure and rapidity of the trial timetable, with limited time between two major
trials (both of which were dealing with significant and complex cross-cutting
issues), and a significant amount of evidence for the HIT falling due during the
CIT, inevitably put me and my team under pressures during the CIT and in the
run-up to the HIT that were extreme even in the context of heavy commercial
litigation.
1010. In relation to the Common Issues, I reflect back on the decision made to oppose
all the Claimants’ implied terms, including duties to act in good faith, which were
at the heart of the CIT. This was done at the time in the belief that even conceding
a few, apparently reasonable, implied terms would open the door to more on
onerous terms. I still believe that was a reasonable strategy (and one I note that
Lord Grabiner and Lord Neuberger endorsed in relation to the CITJ appeal). But,
taking such a rigid legalistic approach gave an impression that POL was
unyielding and perhaps unreasonable, and this may have contributed to Mr
Justice Fraser's overall perception of POL and my firm's conduct of the litigation.
1011. A further consideration which I have touched upon above, relates to our
preparation of the evidence of the Fujitsu witnesses. I have explained that Fujitsu
was the ultimate source of all our information about the Horizon IT system.
Consequently, when it came to factual evidence about the workings of the
Horizon system, the relevant witnesses had to come from Fujitsu; there was no
Page 554 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
other option. At the same time, POL, the Counsel team and I were aware at the
time of the group litigation that reliance on Fujitsu was not without challenges.
1012. The difficulty when it came to the preparation of the HIT witness evidence,
however, was that there was insufficient time both (i) to probe the factual content
of the draft statements in the level of detail that my team would have ideally liked,
and (ii) to reconsider the approach when it became apparent that the witnesses
Fujistsu had held out, even between them, could not speak to all of the Horizon
Issues with level of the detail or authority POL required. This combination of
factors was, in my assessment, the main reason why the factual evidence for the
HIT contained inaccuracies.
1013. With the full benefit of hindsight (knowing how the evidence came out), I would
probably have wanted to start proofing the Fujitsu witnesses earlier, as this may
have revealed gaps and deficiencies in their knowledge at a stage when we may
have still had time to reconsider the approach (for example, by calling a larger
number of witnesses to cover individual topics within the Horizon Issues). Whilst
I might have wanted to start proofing the witnesses earlier, we were constrained
in practice because many points that Fujitsu were ultimately asked to address
did not became clear (or even known) until POL received the Claimants’ evidence
and Mr Coyne's first report in Autumn 2018. We could therefore have started the
process of interviewing potential witnesses sooner, but it is hard to say whether
that would have avoided the later problems encountered with the HIT evidence.
1014. The other point on which I consider in hindsight we might have acted differently
is in relation to the KEL database. I have explained above why, based on the
information we received from Fujitsu, I thought in October 2017 that the KEL was
not likely to contain any information of value to the Claimants’ case, and was
Page 555 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
likely to be disproportionate to search and analyse in detail (see §§553-594
above). Subsequent experience showed that was incorrect, in that analysis of
the KEL by the parties’ experts did reveal evidence of bugs. I naturally look back
and wonder whether we might have pressed Fujitsu harder or earlier for more
detailed information about the KEL. It is, however, impossible for me to say
whether pushing Fujitsu on this point earlier in 2017 would have made a
meaningful difference, as they were clear in their conviction that the KEL could
not be relevant to the Claimants’ claims (and it was the experts, Mr Coyne and
Dr Worden, who showed the contrary to be true). Further, in any event, I do not
believe that this would ultimately have made a meaningful difference to the
overall trajectory of the litigation. Mr Coyne was given access to the KEL at the
time the Court first started making orders for disclosure (and nearly a year before
the HIT began), and his and Dr Worden’s detailed investigations into it followed
quickly thereafter.
1015. Generally, I observe that the group litigation was an adversarial process and
that meant sometimes advising that POL take steps that were adverse to the
interests of some SPMs. Nevertheless, throughout my work for POL, I believe
my firm was consistent in advising POL to accept when it had got things wrong,
and throughout we advised POL to explore settlement options at appropriate
junctures. I believe that our advice was reasonable and in line with my
professional duties as a solicitor. As I have explained, I had no difficulty in giving
firm advice to POL. I had a longstanding professional relationship with POL by
the time of the group litigation. When acting on a matter for so long, it is possible
that some measure of unconscious confirmation bias can influence one's thinking
— that is almost inevitable — but POL never had any compunction about bringing
Page 556 of 557
WITN10390200
WITN10390200
DocuSign Envelope ID: CF2EA3BB-4FE8-4255-8846-1BE4BF70BC90
in fresh advisers to give a second opinion (as they did, for example, with
Linklaters during the Mediation Scheme, Deloitte, Jonathan Swift QC, David
Cavender QC (part-way through the group litigation), Lord Neuberger and Lord
Grabiner QC for the Recusal Application, and also Norton Rose and Herbert
Smith Freehills) and so I do not believe that, if this did happen, it materially
influenced the outcome.
1016. Finally, Q122 of the Request asks me whether there are any other matters I
would like to bring to the Chair’s attention. As I hope is clear from this statement,
I have endeavoured to provide the Inquiry with as much useful information as
possible within the limits of the time available and the very large number of
potentially relevant documents. To that end, I have not strictly confined myself to
the questions set out in the Request but rather have sought to provide the
relevant context to those questions, and my answers, wherever helpful and
practicable to do so. I hope that my additional observations throughout this
statement will be of assistance to the Inquiry in its important task.
Statement of truth
I believe the content of this statement to be true.
Signed:
Andrew Parsons
Date:
Page 557 of 557
WITN10390200
WITN10390200
WOMBLE
BOND
DICKINSON
ANNEX 1
INDEX OF DOCUMENTS REFERENCED IN WITNESS STATEMENT OF ANDREW PAUL PARSONS
No. I Document Reference Document
1. POL00099063 Second Sight's Interim Report dated 07.07.13
2. WBON0000726 Email chain between Ivan Swepson, Rodric Williams, Gavin Matthews and Andrew
Parsons dated 05.04.13 — 08.04.13
3. POL00098035 Email from Andrew Parsons to Simon Baker, Lin Norbury, Gareth Jenkins, Dave
Posnett and Rod Ismay dated 19.04.13
4. POL00098294 Email chain between Ron Warmington, Simon Baker, Alwen Lyons, Alan Bates, Kay
Linnell, Susan Crichton, Andrew Parsons, Andrew Winn, Craig Tuthill, Dave Posnett,
Gareth Jenkins, lvan Swepson, Lin Norbury, Rod Ismay and Rodric Williams dated
11.05.13 — 17.05.13
5. WBON0000736 Email exchange between Andrew Parsons, Rodric Williams and Simon Baker dated
22.05.13
6. WBON0000343, Email chain between lan Henderson, Alwen Lyons, Susan Crichton, Steve Allchorm,
Ron Warmington, Rosie Gaisford and Andrew Parsons dated 24.05.13- 10.06.13
7. WBON0000344 Horizon Spot Review 5 Response - Centrally Input Transactions undated
8. POL00130311 Email chain between Simon Baker, Steve Allchorn and Andrew Parsons dated
12.06.13 - 19.06.13
9. POL00188299 Notes from Second Sight / Post Office Meeting. Attendees: Ron Warmington, Pete
Newsome, Steve Alichom, Alwen Lyons, Lesley Sewell and James Davidson dated
12.06.13
10. POL00098619 Email chain between Alan Bates, Ron Warmington, Simon Baker, Steve Alichorn, Pete
Newsome, Gareth Jenkins, Rod Ismay and Andrew Parsons dated 13.06.13 —
20.06.13
11. WBON0000739 Email chain between lan Henderson, Alwen Lyons, Susan Crichton, Steve Allchom
and Andrew Parsons dated 24.05.13- 20.06.13
12. POL00031348 Email chain between Simon Baker, Steve Allchorn and Andrew Parsons dated
12.06.13 — 19.06.13
13. POL00031350 Email chain between Ron Warmington, Rod Ismay, Angela Van Den Bogerd, Steve
Allchorn and Andrew Parsons dated 05.02.13 — 20.06.13
14. WBON0000737 Email chain between Simon Baker, Andrew Parsons and Steve Allchorn dated
13.06.13 — 18.06.13.
15. WBON0000361 Email from Andrew Parsons to Steve Allchorn dated 20.06.13
16. WBON0000363 Draft Horizon Spot Review 5 — Response - undated
17. WBON0000389 Email chain between Steve Allchorn, Ron Warmington and lan Henderson dated
21.06.13 - 24.06.13.
AC_207339548_5
WITN10390200
WITN10390200
No. I Document Reference Document
18. POL00243412 Horizon Spot Review 5 Response -— Bracknell Site & Centrally Input Transactions
undated
19. WBON0000743 Email exchange between Andrew Parsons, Rodric Williams and Andrew Pheasant
dated 01.07.13
20. POL00296872 Email chain including Michael Rudkin, Ron Warmington, Simon Baker and Rodric
Williams dated 01.07.13
21. WBON0000919 Email chain between Simon Baker, Andrew Parsons and Ron Warmington dated
01.08.13 - 05.08.13
22. WBON0000366 Email from Rosie Gaisford to Andrew Parsons dated 05.07.13
23. POL00021822 Email from Rodric Williams to Andrew Parsons and Hugh Flemington dated 27.06.13
24. POL00021823 Raising Concerns with Horizon dated 17.12.12
25. POL00021824 Raising Concerns with Horizon dated 17.12.12
26. WBONO000741 Email chain between Andrew Parsons and Rodric Williams dated 27.06.13
27. WBON0000364 Email chain between Andrew Parson and Rodric Williams dated 27.06.13
28. WBONO0000365 Draft Response to Second Sight Interim Report- undated
29. WBONO0000742 Email from Rodric Williams to Andrew Parsons dated 28.06.13
30. POL00341337 Email chain between Rodric Williams, Andrew Parsons and Andrew Winn dated
28.06.13
31 WBONO0000131 Email chain between Simon Baker, Rodric Williams, Gavin Matthews, Andrew Parsons
and Andrew Pheasant dated 28.06.13- 01.07.13.
32. POL00407496 Email chain between Simon Baker, Rodric Williams, Gavin Matthews and Andrew
Parsons dated 28.06.13 — 01.07.13.
33. WBON0000746 Email chain between Simon Baker, Rodric Williams, Gavin Matthews and Andrew
Parsons dated 28.06.13 — 01.07.13
34, WBONO0000757 Email chain between Mark Davies, Lesley Sewell, Alwen Lyons, Susan Crichton,
Rodric Williams, Hugh Flemington, Nina Arnott, Ruth Barker, Simon Baker and Andrew
Parsons dated 03.07.13
35. POL00190547 Draft Letter to James Arbuthnot MP undated.
36. WBON0000759 Email chain between Andrew Parsons and Rodric Williams dated 04.07.13
37. WBON0000135 Email exchange between Andrew Parsons, Rodric Williams and Hugh Flemington
dated 04.07.13
38. WBON0000136 Draft letter to James Arbuthnot MP
39. POL00021745 Email chain between lan Henderson, Simon Baker, Mark Davies, Martin Edwards,
Lesley Sewell, Susan Crichton, Alwen Lyons, Angela Van Den Bogerd, Rodric
AC_207339548_5
WITN10390200
WITN10390200
No. I Document Reference Document
Williams, Hugh Flemington, Nina Arnott, Ruth Barker, Paula Vennells and Andrew
Parsons dated 05.07.13
40. WBON0000760 Email chain between lan Henderson, Simon Baker, Mark Davies, Martin Edwards,
Lesley Sewell, Susan Crichton, Alwen Lyons, Angela Van Den Bogerd, Rodric
Williams, Hugh Flemington, Nina Arnott, Ruth Barker, Paula Vennells and Andrew
Parsons dated 05.07.13
41. WBON0000762 Email chain between lan Henderson, Simon Baker, Mark Davies, Martin Edwards,
Lesley Sewell, Susan Crichton, Alwen Lyons, Angela Van Den Bogerd, Rodric
Williams, Hugh Flemington, Nina Arnott, Ruth Barker, Paula Vennells and Andrew
Parsons dated 05.07.13
42. WBON0000134 Email chain involving Andrew Parsons, Andrew Pheasant and Rodric Williams dated
05.07.13
43. WBON0000763 Email chain between lan Henderson, Susan Crichton, Hugh Flemington, Rodric
Williams and Andrew Parsons dated 07.07.13
44, POL00039991 Email chain between Martin Edwards, Paula Vennells, Mark Davies, Alwen Lyons,
Lesley Sewell, Susan Crichton, Hugh Flemington, Rodric Williams, Ruth Barker, Nina
Arnott and Andrew Parsons dated 08.07.13
45. WBONO000766 Email from Susan Crichton to Andrew Parsons dated 12.07.13
46. WBONO0000767 Email chain between Susan Crichton, Andrew Parsons and Rodric Williams dated
12.07.13
47. POL00230639 Email chain between Susan Crichton, Andrew Parsons and Rodric Williams dated
12.07.13
48. POL00407537 Email chain between Susan Crichton, Rodric Williams, Andrew Parsons and Alwen
Lyons dated 12.07.13
49. WBON0000768 Email chain between Alan Bates, Paula Vennells, Alwen Lyons, Susan Crichton,
Rodric Williams and Andrew Parsons dated 11.07.13 12.07.13
50. WBON0000769 Email chain between Alan Bates, Paula Vennells, Alwen Lyons, Susan Crichton,
Rodric Williams, Andrew Parsons and Hugh Flemington dated 11.07.13 - 12.07.13
51. POL00407548 Email chain between Susan Crichton and Gavin Matthews dated 17.07.13— 18.07.13
52. WBON0000775 Email chain between Susan Crichton, Andrew Parsons and Mark Davies dated
17.07.13 — 19.07.13
53. WBON0000776 Email chain between Simon Baker, Martin Edwards, Alwen Lyons, Mark Davies, Susan
Crichton and Andrew Parsons dated 17.07.13 19.07.13
54. POL00117035 Mediation Proposal
55. POL00117034 Email chain between Andrew Parsons, Mark Davies, Susan Crichton, Alwen Lyons,
Gavin Matthews, Simon Richardson, Hugh Flemington, Rodric Williams and Jarnail A
Singh dated 19.07.13- 21.07.13
56. POL00192226 Email chain between Simon Baker and Ron Warmington dated 16.07.13 — 17.07.13
AC_207339548_5
WITN10390200
WITN10390200
No. I Document Reference Document
57. POL00099445 Email chain between Peter Batten, Susan Crichton, Martin Edwards and Andrew
Parsons dated 12.08.13 — 19.08.13
58. WBON0000778 Email between Simon Baker and lan Henderson, Ron Warmington, Susan Crichton,
Andy Holt, Alwen Lyones, Andrew Parsons, Angela Van Den Bogerd, Mark R Davies,
Alan Bates and Ruth Barker dated 24.07.13
59. WBONO0000784 Email between Simon Baker, lan Henderson, Ron Warmington, Susan Crichton, Andy
Holt, Alwen Lyons, Andrew Parsons, Angela Van Den Bogerd, Mark R Davies, Alan
Bates, Kay Linnell and Ruth Barker dated 29.07.13
60. WBON0000787 Email between Simon Baker, lan Henderson, Ron Warmington, Susan Crichton, Andy
Holt, Alwen Lyons, Andrew Parsons, Angela Van Den Bogerd, Mark R Davies, Alan
Bates, Kay Linnel, Rodric Williams and Sophie Bialaszewski dated 06.08.13
61. WBON0000790 Email between Simon Baker, Andrew Parsons, Susan Crichton and Rodric Williams
dated 12.08.13
62. WBONO0000817 Terms of Reference For Working Group
63. WBONO0000789 Terms of Reference for Independent Expert
64. WBONO0000795 Email chain between Louise Kelly, Andrew Parsons and Rodric Williams dated
20.08.13
65. WBON0000773 Email from Tracy Hunter to Andrew Parsons dated 18.07.13
66. POL00022598 Helen Rose Report 12 June 2013
67. WBONO0000751 Email chain between Dave Posnett, Rodric Williams and Andrew Parsons dated
14.06.13 — 03.07.13
68. POL00186743 Horizon Spot Review Response — SRO1: Debt Cards — Cash Withdrawals and GIRO
Payments undated.
69. POL00193002 Prosecutions' Expert Evidence — Advice on the Use of Expert Evidence relating to the
Integrity of the Fujitsu Services Ltd Horizon System dated 15.07.13
70. WBON0000770 Email chain between Martin Smith, Susan Crichton, Rodric Williams and Andrew
Parsons dated 17.07.13
71. WBONO0000765 Email chain between Hugh Flemington, Simon Clarke, Martin Smith and Andrew
Parsons dated 08.07.13
72. POL00039996 Email from Susan Crichton to Andrew Parsons, Simon Richardson, Gavin Matthews,
Hugh Flemington, Rodric Williams and Jarnail Singh dated 16.07.13
73. POL00039993 CCRC Reviewing Criteria
74, POL00039996 Email from Susan Crichton to Andrew Parsons dated 16.07.13
75. POL00407546 Email chain between Susan Crichton, Andrew Parsons and Gavin Matthews dated
16.07.13
76. WBON0000133 Email chain between Susan Crichton, Andrew Parsons and Gavin Matthews dated
16.07.13 - 17.07.13
AC_207339548_5
WITN10390200
WITN10390200
No. I Document Reference Document
77. WBONO0000777 Email chain Rodric Williams, Susan Crichton, Andrew Parsons and Jarnail Singh dated
22.07.13 — 23.07.13
78. WBON0000782 Email from Gavin Matthews to Susan Crichton, Hugh Flemington, Jarnail Singh and
Rodric Williams dated 26.07.13
79. POL00297983 Draft Letter from Susan Crichton to S Berlin (CCRC) dated 26.07.13
80. WBONO0001705 Email Jarnail Singh to Simon Clarke and Andrew Parsons dated 23.10.13
81. WBON0000806 Email chain between Andrew Parsons, Martin Smith, Rodric Williams and Harry
Bowyer dated 05.08.13 — 06.08.13
82. POL00145832 Draft Mediation Pack undated (circulated on 06.08.13)
83. POL00083932 Minutes of Horizon Regular Call dated 19.07.13
84, WBON0000772 Email from Ben Thorp to Andrew Parsons dated 18.07.13
85. POL00083933 Minutes of Horizon Regular Call dated 24.07.13
86. POL00083931 Minutes of Horizon Regular Call dated 07.08.13
87. POL00083930 Minutes of Horizon Regular Call dated 14.08.13
88. POLO00193767 Notes from Call regarding Horizon Issues. Attendees: Rodric Williams, Martin Smith,
Jamail Singh, Rod Ismay, Dave Posnett, Rob King, Sophie Bialaszewski, Gayle
Peacock, Steve Beddoe and Kendra Dickinson dated 31.07.13
89. POL00139691 Email chain between Dave Posnett, Andrew Parsons, Rob King and Jarnail Singh
dated 16.08.13 — 19.08.13
90. POL00193596 Email chain between Dave Posnett, Andrew Parsons, Rob King and Jarnail Singh
dated 16.08.13 - 19.08.13
91. WBON0001710 Email chain involving Andrew Parsons, Martin Smith, Susan Crichton, Gavin Matthews
and Simon Richardson dated 13.08.13 to 15.08.13
92. WBONO0000791 Email chain between Martin Smith, Susan Crichton and Andrew Parsons dated
13.08.13
93. POL00229411 Advice Note on Disclosure — The Duty to Record and Maintain Material dated 02.08.13
94. POL00139693 Email chain between Jarnail Singh, John Scott, Rob King, Martin Smith, Andrew
Parsons, Rodric Williams, Susan Crichton, Hugh Flemington, Kayleigh-Lee Harding
and Gayle Peacock dated 20.08.13 — 22.08.13
95. POL00137427 Horizon Weekly Report
96. WBONO0000796 Email from Andrew Parsons to Dave Posnett, Rodric Williams, Jarnail Singh, Martin
Smith, Rod Ismay, Andrew Winn, Rob King, Sophie Bialaszewski, Ruth Barker, Nick
Beal, Gayle Peacock, Steve Beddoe, Jeff Burke, Kayleigh-Lee Harding and Kendra
Dickinson dated 21.08.13
97. WBONO0000807 Email chain between Hugh Flemington, Rob King and Andrew Parsons dated 22.08.13
AC_207339548_5
WITN10390200
WITN10390200
No. I Document Reference Document
98. POL00139696 Protocol to Horizon Regular Calls
99. POL00139695 Email from Gayle Peacock to Dave Posnett, Jeff Burke, Sophie Bialaszewski, Anne
Allaker, Jarnail Singh, Rod Ismay, Rodric Williams, Rob King, Andrew Winn, Nick Beal,
Kendra Dickinson, Steve Beddoe, Ruth Barker, Kathryn Alexander, Shirley Hailstones,
Martin Smith, Andrew Parsons, Andy Hayward and Kayleigh-Lee Harding dated
09.10.13
100. I POLO0006485 Notes from Conference with Brian Altman QC. Attendees: Brian Altman QC, Susan
Crichton, Rodric Williams, Jarnail Singh, Simon Clarke, Harry Bowyer, Martin Smith,
Gavin Matthews and Andrew Pearson dated 09.09.13
101. I POL00223376 Interim Review of Cartwright King's Current Process dated 02.08.13
102. I WBON0000393 Email chain between Brian Altman, Gavin Matthews, Simon Richardson, Susan
Crichton, Jarnail Singh, Rodric Williams, Piero D'Agostino and Andrew Parsons dated
02.08.13 — 04.08.13
103. I WBONO0000786 Email chain between Brian Altman, Gavin Matthews, Simon Richardson, Susan
Crichton, Jarnail Singh, Rodric Williams and Piero D'Agostino dated 02.08.13—
05.08.13
104. I POL00229413 Response to the Interim Review of Cartwright King's Current Process by Brian Altman
QC dated 13.08.13
105. I POL00139866 Notes from Conference with Brian Altman QC. Attendees: Brian Altman QC, Susan
Crichton, Rodric Williams, Jarnail Singh, Simon Clarke, Harry Bowyer, Martin Smith,
Gavin Matthews and Andrew Pearson dated 09.09.13
106. I POL00333840 Email chain between Brian Altman and Gavin Matthews dated 20.09.13— 23.09.13
107. I WBONO000725 Handwritten Note of Meeting. Attendees: Martin Smith, Andrew Bolc, Simon Clarke,
Andrew Parsons, Jarnail Singh, Rodric Williams, Gavin Matthews and Brian Altman
dated 04.10.13
108. I WBON0001702 Draft Settlement Policy dated December 2013
109. I POL00201950 Email chain involving Andrew Parsons, Angela Van Den Bogerd, Belinda Crowe, Rodric
Williams, and Ron Warmington dated 17.03.14 — 19.03.2014
110. I POL00021860 Email chain involving Kendra Dickinson, Claire Parmenter, Andrew Parsons, Angela
Van Den Bogerd, Sue Richardson, Rod Ismay, Sophie Bialaszewski, Gayle A Peacock
and Belinda Crowe dated 20.12.13 — 27.12.13
111. I POL00026656 Face to Face meeting of the Working Group -— Initial Complaint Review and Mediation
Scheme dated 7.03.14
112. I WBONO0000824 Minute — Initial Complaint and Mediation Scheme Working Group dated 1.04.14
113. I POL00040066 Post Office Limited Initial Complaint Review and Mediation Scheme Overview of
Horizon and branch trading practices
114. I WBON0000396 Email from Claire Parmenter to Andrew Parsons dated 10.01.14
115. I WBON0000398 Email from Claire Parmenter to Andrew Parsons dated 11.02.14
AC_207339548_5
WITN10390200
WITN10390200
No. I Document Reference Document
116. I WBON0000401 Draft Overview of Horizon and branch trading practices
117. I WBONO0000402 Email chain involving Rod Ismay, Claire Parmenter, Alison Bolsover, Andrew Winn,
Lorraine Garvey and Kay Wilson dated 30.01.14 —- 11.02.14
118. I WBONO000812 Email from Andrew Parson to Nicky Mal and Belinda Crowe dated 21.02.14
119. I WBONO000813 Draft - Post Office Limited - Initial Complaint Review and Mediation Scheme -
Overview of Horizon and branch trading practices undated
120. I WBONO000814 Email chain involving Claire Parmenter, Nicky Mal, Andrew Parsons, and Belinda Crowe
dated 21.02.14
121. I POL00026666 Working Group for the Initial Complaint Review and Mediation Scheme Key points and
actions from the conference call at 1pm on 12 December 2013
122. I WBONO000808 Email chain involving Andrew Pheasant, Zoe Topham, Alison Bolsover, Rodric
Williams, Rebecca Butler, Stacey J Beresford, Jenny Smith, Darryl Webb and Pat
Davies dated 13.12.11 — 26.11.13
123. I WBONO000951 Notice of Issue dated 28.11.13
124. I WBONOO000809 Letter from WBD to Mr T Walters dated 13.12.13
125. I WBONO001670 Email from Andrew Parsons to Ron Warmington, lan Henderson, Alan Bates, Anthony
Hooper, Kay Linnell, Belinda Crowe and Angela Van-Den-Bogerd dated 13.12.13
126. I WBONOO00950 Consent Order dated 29.01.13
127. I WBONOO000890 Letter from WBD to The Court Manager dated 06.08.14
128. WBONO000949 General Form of Judgment or Order dated 09.09.14
129. I WBON0001667 Consent Order dated 13.10.20
130. I POL00026643 Working Group for the Initial Complaint Review and Case Mediation Scheme Standing
Agenda for Thursday Calls dated 13.03.14
131. I POL00026672 Minute - Working Group for the Initial Complaint Review and Case Mediation Scheme
dated 10.07.14
132. I POL00026652 Working Group for the Initial Complaint Review and Case Mediation Scheme — Minute
dated 17.04.14
133. I WBONO0000821 Email chain involving Rodric Williams, Andrew Parsons, David Oliver, Angela Van-
Den-Bogerd, Ron Warmington and Chris Holyoak dated 26.03.14 — 27.03.14
134. I WBON0000822 Email chain involving Shirley Hailstones, Rodric Williams, Angela Van-Den-Bogerd,
Chris Aujard, Andrew Parsons, Kathryn Alexander, lan Henderson and Ron
Warmington dated 28.11.13- 29.11.13
135. I WBONO000404 Email from Andrew Parsons to Ben Thorp, Thomas Lillie, Paul Loraine, Alexandra
Ward, Claire Parmenter, Alva Leigh-Doyle, Andrew Pheasant, Richard Pike and
Matthew Harris dated 21.05.14
AC_207339548_5 7
WITN10390200
WITN10390200
No. I Document Reference Document
136. I WBONO0000820 Email chain involving Andrew Parsons, Angela Van-Den-Bogerd, David Oliver, Chris
Aujard, Ron Warmington, lan Henderson and Chris Holyoak dated 26.03.14
137. I WBONO000847 Email exchange between David Oliver, Andrew Parsons, Chris Aujard, Rodric
Williams, Belinda Crowe and Angela Van-Den-Bogerd dated 28.04.14 — 29.04.14
138. I POL00220159 Complaint Review and Mediation Scheme A paper prepared by Post Office to assist
Second Sight with the finalisation of their Briefing Report— Part Two Version two
undated
139. I WBON0000413 Horizon Tracker Spreadsheet - 13.05.15
140. I POL00021814 Email exchange between Andrew Parsons, Belinda Crowe, Chris Aujard, Angela Van-
Den-Bogerd, Jonathan Swil and Rodric Williams dated 01.08.14 - 06.08.14
141. I POLO0207175 Draft Post Office Response to Second Sight's Draft Part 2 Report- undated
142. I POL00074462 Email from Andrew Parsons to Chris Aujard, Rodric Williams, David Oliver, Belinda
Crowe, Sophie Bialaszewski and Angela Van Den Bogerd dated 6 March 2014
143. I WBONO000853 Email chain involving Belinda Crowe, Amanda A Brown, lan Henderson and Ron
Warmington dated 09.05.14 — 22.05.14
144. I POL00026662 Minute — Initial Complaint Review and Mediation Scheme Working Group dated
20.05.14
145. I POL00116487 Email chain involving Andrew Parsons, Belinda Crowe, Amanda A Brown, Priti Singh
and Chris Aujard dated 07.04.14 - 08.04.14
146. I POL00129392 Email chain involving Allison Drake, Shirley Hailstones, Jane M Owen, Joanne
Hancock, Paul J Smith, Jim Coney, Keith Scott, Peter Todd, Wayne Z Griffiths, Wendy
Mahoney, Shirley Hailstones, Jarnail Singh, Andrew Parsons, Chris Aujard, Rodric
Williams, Angela Van-Den-Bogerd, Belinda Crowe, David Oliver and Kathryn
Alexander dated 17.06.14 — 16.07.14
147. I WBONO000888 Email chain involving Andrew Pheasant, Peter Todd, Shirley Hailstones and Matthew
Harris dated 25.07.14 — 31.07.14
148. I WBONO000889 Initial Complaint Review and Mediation Scheme Issues analysis dated 24.07.17
149. I WBONO0000828 Email chain involving Andrew Parsons, Simon Clarke, Rodric Williams and Dave
Posnett dated 14.06.13 — 09.04.14
150. I WBONO0000834 Email chain involving Simon Clarke, Andrew Parsons, Rodric Williams and Dave
Posnett dated 14.06.13— 09.04.14
151. I WBONO0000838 Email chain involving Andrew Parsons, Belinda Crowe, Angela Van-Den-Bogerd, Andy
Holt, Rodric Williams, Steve Darlington and Ron Warmington dated 08.04.14—
14.04.14
152. I WBONO0000825 Email chain involving Jarnail Singh, Andrew Parsons, Rodric Williams, Jonny Gribben,
Harry Bowyer and Andrew Bolc dated 07.04.14 - 08.04.14
153. I POL00061369 Email exchange between Jarnail Singh, Andrew Parsons, Angela Van-Den-Bogerd,
Rodric Williams, Kathryn Alexander, Shirley Hailstones, dated 17.04.14 - 23.04.14
AC_207339548_5 8
WITN10390200
WITN10390200
No. I Document Reference Document
154. I WBON0000443 Email exchange between Andrew Parsons, and Brian Altman dated 18.07.16—
22.07.16
155. I WBONO000848 Email exchange between Shirley Hailstones, Andrew Parsons and Angela Van-Den-
Bogerd dated 27.04.14 — 06.05.14
156. I WBONO000849 Email chain involving Andrew Parsons, Shirley Hailstones, Angela Van-Den-Bogerd
and Rodric Williams dated 27.04.14 — 06.05.14
157. I POL00046216 Email from Andrew Parsons to Martin Smith dated 06.05.14
158. I POL00046219 Email exchange between Martin Smith and Andrew Parsons dated 07.05.14- 08.05.14
159. I WBONO000850 Email chain involving Andrew Parsons, Jarnail Singh, Chris Aujard, Jessica Madron
and Martin Smith dated 07.05.14 - 08.05.14
160. I WBONO0000403 Email chain involving Rodric Williams, Jarnail Singh, Andrew Parsons, Chris Aujard,
Jessica Madron and Martin Smith dated 07.05.14 - 08.05.14
161. I WBONO000851 Email chain involving Andrew Parsons, Shirley Hailstones, Kathryn Alexander, Angela
Van-Den-Bogerd, Rodric Williams, Jarnail Singh, Chris Aujard, Jessica Madron and
Martin Smith dated 07.05.14 - 09.05.14
162. FUJ00087119 Email chain involving Sean Hodgkinson, James Davidson, Mark Westbrook, Rod
Ismay, Pete Newsome, Torstein Godeseth, Bill Membery, Rodric Williams, Andrew
Parsons and Michael Harvey dated 14.04.14 - 19.05.14
163. I POL00117650 Email exchange between Andrew Winn and Alan Lusher dated 15.10.08— 23.10.08
164. I WBON0000826 Email chain involving Andrew Parsons, Belinda Crowe, Angela Van-Den-Bogerd, Andy
Holt, Rodric Williams, Steve Darlington and Ron Warmington dated 08.04.14
165. I WBONO0000827 Email chain involving Angela Van-Den-Bogerd, Belinda Crowe, Andrew Parsons, Andy
Holt, Rodric Williams, Ron Warmington and Steve Darlington dated 08.04.14 —
09.04.14
166. I WBONO000835 Email chain involving Angela Van-Den-Bogerd, Belinda Crowe, Andrew Parsons, Andy
Holt, Rodric Williams, Ron Warmington and Steve Darlington dated 08.04.14—
14.04.14
167. I WBONO000837 Email exchange between Rodric Williams and Andrew Parsons dated 14.04.14
168. I WBONO000845 Email chain involving Andrew Parsons, Rodric Williams and James Davidson dated
14.04.14 - 22.04.14
169. I POL00204068 Initial Complaint Review and Mediation Scheme Horizon Data document dated
22.04.14
170. I WBONO000852 Email chain involving Andrew Parsons, Angela Van-Den-Bogerd, Rodric Williams and
James Davidson dated 14.04.14 - 09.05.14
171. I WBONO000854 Email chain involving Andrew Parsons, Rodric Williams and James Davidson dated
14.04.14 — 28.05.14
172. I WBONO000860 Email from Andrew Parsons to David Oliver dated 09.06.14
AC_207339548_5 9
WITN10390200
WITN10390200
No. I Document Reference Document
173. I POLO0307712 Initial Complaint Review and Mediation Scheme — M056 Post Office Investigation
Report - undated
174. I POL00028062 Horizon: Desktop Review of Assurance Sources and Key Control Features Draft for
discussion dated 23.05.14
175. I WBONO000891 Email exchange between Mark Westbrook, Rodric Williams, Andrew Parsons and
Gareth James dated 21.07.14 — 22.08.14
176. I WBONO000856 Email chain involving Andrew Parsons, Rodric Williams, Gareth James, Chris Aujard
and Julie George dated 29.05.14 - 30.05.14
177. I WBONOOO00960 Email exchange between Rodric Williams and Andrew Parsons dated 10.02.16—
08.03.16
178. I POL00105635 Deloitte Project Zebra — Phase 1 Report - For validation in advance of Board
discussion on Wednesday 30 April
179. I POL00031384 Deloitte HNG-X: Review of Assurance Sources — Discussion Areas re: Phase 2 -
undated
180. I POL00031391 Deloitte HNG-X: Review of Assurance Sources Phase 1— Board Update dated
13.05.14
181. I POL00029726 Deloitte HNG-X: Review of Assurance Sources — Board Update dated 16.05.14
182. I POL00226961 Initial Complaint Review and Mediation Scheme Briefing Report— Part Two Prepared
by Second Sight dated 21.08.14
183. I WBONO000908 Email chain involving Andrew Parsons, Rodric Williams, Belinda Crowe, James
Davidson, Tom Wechsler, Melanie Corfield, Angela Van-Den-Bogerd, Patrick Bourke,
Mark Underwood and Andy Holt dated 21.10.14
184. I POLOO211255 Initial Complaint Review and Mediation Scheme Horizon Data document - undated
185. I WBONO000916 Telephone attendance note between Mark Westbrook from Deloitte and Andy Parsons
dated 03.11.14
186 WBON0000910 Email exchange between Mark Westbrook, Rodric Williams and Andrew Parsons
dated 04.11.14- 10.11.14
187. I WBONOO00911 Email chain involving Sean Hodgkinson, Mark Westbrook, James Davidson, Julie
George, John Simpkins, Jane E Smith, Rod Ismay and Dave M King dated 13.05.14—
15.05.14
188. I WBONO000912 Email from Andrew Parsons to Belinda Crowe, Patrick Bourke and Andy Holt dated
10.11.14
189. I POLO0212054 Draft Initial Complaint Review and Mediation Scheme Horizon Data document -
undated
190. I WBONOO000914 Email exchange between Patrick Bourke, Andrew Parsons, Belinda Crowe and Andy
Holt dated 10.11.14
191. I WBONO0000479 Email chain involving Amy Prime, Andrew Parsons and Owain Draper dated 05.04.17
— 06.04.17
AC_207339548_5 10
WITN10390200
WITN10390200
No. I Document Reference Document
192. I WBONO000917 Email exchange between Mark Underwood and Andrew Parsons dated 10.12.14—
15.12.14
193. I WBON0000327 Draft Initial Complaint Review and Mediation Scheme Horizon Data document -
undated
194. I POL00408247 Email exchange between Andrew Parsons, Belinda Crowe, Angela Van-Den-Bogerd
and Rodric Williams dated 28.02.15 — 02.03.15
195. I POLO0021845 Email chain involving Patrick Bourke, Andrew Parsons, lan Henderson, Jane
MacLeod, Ron Warmington and Chris Holyoak dated 02.04.15- 07.04.15
196. I POLO0225912 Document entitled Transactions not entered by the Sub-Postmaster or their staff -
undated
197. I POL00225913 Receipts/Payments Mismatch issue notes - undated
198. I POL00225914 Document prepared by Gareth Jenkins entitled Correcting Accounts for "lost"
Discrepancies - dated 29.09.10
199. I WBONO000924 Email from Mark Underwood to Kevin Lenihan, James Davidson and Newsome Pete
dated 07.04.15
200. I POL00243542 Draft Complaint Review and Mediation Scheme Horizon Data document - undated
201. I WBONO000927 Email chain involving Pete Newsome, Mark Underwood and Andrew Parsons dated
08.04.15
202. I WBONO000928 Email chain involving Patrick Bourke, Mark Underwood, Pete Newsome and Andrew
Parsons dated 08.04.15
203. I WBONO0000929 Email chain involving Pete Newsome, Mark Underwood and Andrew Parsons dated
08.04.15
204. I POL00041040 Email chain involving Andrew Parsons, Partick Bourke, Mark Underwood and Pete
Newsome dated 08.04.15
205. I POL00226089 Draft response to SS - undated
206. I POLO0021785 Email chain involving Mark Underwood, Andrew Parsons, Partick Bourke, lan
Henderson, Jane MacLeod, Ron Warmington and Chris Holyoak dated 02.04.15-
08.04.15
207. I WBONO000930 Email chain involving Michael Harvey, Mark Underwood and Pete Newsome dated
09.04.15 — 10.04.15
208. I WBON0000922 Email from Simon Clarke to Rodric Williams and Andrew Parsons dated 27.03.15
209. I POL00228075 Cartwright King Note: Deloitte Report- Questions for POL - undated
210. I WBON0000340 Email from Andrew Parsons to Rodric Williams dated 27.03.15
211. I WBONO0000931 Email from Andrew Parsons to Simon Clarke and Martin Smith dated 05.05.15
212. I WBONO0000944 Email from Andrew Parsons to Martin Smith and Simon Clarke dated 15.07.15
AC_207339548_5 "1
WITN10390200
WITN10390200
No. I Document Reference Document
213. I WBONO000946 Email from Andrew Parsons to Brian Altman dated 20.07.15
214. I WBONO000942 Email chain involving Andrew Parsons, Rodric Williams, Mark Underwood, Gavin
Matthews and Pete Newsome dated 26.06.15 — 08.07.15
215. I POL00238791 Old Horizon note Prepared by Gareth Jenkins dated 08.07.15
216. I POL00021775 Email from Andrew Parsons to Rodric Williams dated 21.07.15
217. I POL00021777 Email exchange between Brian Altman and Andrew Parsons dated 27.0815- 28.08.15
218. I POL00026668 Working Group for the Initial Complaint Review and Case Mediation Scheme Working
Group Minute dated 05.06.14
219. I WBONO000132 Initial Complaint Review and Mediation Scheme — M054 Post Office Preliminary
Investigation Report - undated
220. I POL00306593 Post Office Mediation Scheme — Second Sight - M054 Case Review Report dated
11.06.14
221. I WBONO0000859 Settlement Analysis M054 — undated
222. I POL00026664 Working Group for the Initial Complaint Review and Case Mediation Scheme Working
Group Minute dated 12.06
223. I POL00026673 Minute — Initial Complaint Review and Mediation Scheme Working Group dated
16.06.14
224. I WBONO000864 M054 Decision of Sir Anthony Hooper dated 24.06.14
225. I POL00026671 Working Group for the Initial Complaint Review and Case Mediation Scheme Minute of
the Working Group Call dated 17.07.14
226. I WBONO000876 Email chain involving Andrew Parsons, Angela Var-Den-Bogerd and Belinda Crowe
dated 28.07.14
227. I WBON0000877 Email chain involving Chris Aujard, Belinda Crowe, David Oliver, Andrew Parsons and
Alan Bates dated 23.07.14
228. I POL00206822 Extracts From Hansard dated 09.07.13
229. I POL00206823 Overview of the Initial Complaint Review and Mediation Scheme document with
comments - undated
230. I WBONO000874 Initial Complaint Review and Mediation Scheme — Test for Mediation — Post Office
submission
231. I POL00207229 Letter From Sir Anthony Hooper to Chris Aujard re "The mediation test" dated 08.08.14
232. I WBONO000885 Email from Andrew Parsons to Belinda Crowe and Chris Aujard dated 29.07.14
233. I WBON0000886 Post Office Submission on the test for mediation — undated
234. I POL00207393 Email from Andrew Parsons to Belinda Crowe, David Oliver and Angela Van-Den-
Bogerd dated 13.08.14
AC_207339548_5 12
WITN10390200
WITN10390200
No. I Document Reference Document
235. I POL00207394 Post Office submission on the role of the Working Group in Mediating Cases -
undated
236. I WBONO0000893 Email from Belinda Crowe to Kay Linnrll, Anthony Hooper, lan Henderson, Angela
Van-Den-Bogerd, Andrew Parsons, Alan Bates, Rodric Williams, Chris Aujard, Ron
Warmington and Mediation@2ndsight dated 26.08.14
237. I POL00210134 Decision of Sir Anthony Hooper on the Role of the Working Group in deciding whether
cases are suitable for mediation and the test for mediation - undated
238. I WBONO000895 Email chain involving Andrew Parsons, Belinda Crowe, Rodric Williams, David Oliver,
Angela Van-Den-Bogerd, Anthony Hooper, lan Henderson, Alan Bates,
Mediation@2ndsight, Chris Aujard and Ron Warmington dated 26.08.14
239. I POL00210056 Briefing Note Working Group Call dated 02.10.14
240. I POLO0211024 Cases for Discussion at F2F dated 17.10.14
241. I WBONO000900 Email from Rodric Williams to Belinda Crowe, Chris Aujard, Andrew Parsons, David
Oliver and Angela Van-Den-Bogerd dated 15.09.14
242. I WBONOO000902 Email chain involving Stephen Hocking, Rodric Williams, Belinda Crowe and Andrew
Parsons dated 09.09.14
243. I WBON0000867 Email from Martin Smith to Andrew Parsons dated 09.07.14
244. I POL00305248 Advice from Simon Clarke to Post Office Limited dated 09.07.14
245. I WBONO000869 Email chain involving Andrew Parsons, David Oliver and Martin Smith dated 09.07.14
246. I WBONO000870 Email exchange between Andrew Parsons and Martin Smith dated 09.07.14—
14.07.14
247. I WBONO0000871 Email chain involving Gavin Matthews, Bian Altman QC, Andrew Parsons, David
Oliver, Rodric Williams and Martin Smith dated 09.07.14 — 15.07.14
248. I WBON0000406 Email chain involving Gavin Matthews, Jess Webb, Andrew Parsons, David Oliver,
Rodric Williams and Martin Smith dated 09.07.14 — 16.07.14
249. I POLO0214992 Brian Altman QC's Advice on Suggested Approach to Criminal Case Mediation dated
05.09.14
250. I WBONOO000906 Document setting out Sir Anthony Hooper's exercise of casting vote on whether case
M030 suitable for mediation
251. I WBONO000903 Email chain involving Rodric Williams, Andrew Parsons, Simon Clarke, Jarnail Singh,
Patrick Bourke and Martin Smith dated 03.10.14 — 06.10.14
252. I WBONO000905 Email chain involving Rodric Williams, Angela Van-Den-Bogerd, Andrew Parsons,
Tom Wechsler, Anthony Hooper, Belinda Crowe and lan Henderson dated 06.10.14—
08.10.14
253. I WBONO000907 Email chain involving Belinda Crowe, Patrick Bourke, Andrew Parsons, Anthony
Hooper, Rodric Williams, lan Henderson, Ron Warmington, Angela Van-Den-Bogerd
and Tom Wechsler dated 02.10.14 — 15.10.14
AC_207339548_5 13
WITN10390200
WITN10390200
No. I Document Reference Document
254. I WBONO0000407 Email chain involving Rodric Williams, Melanie Corfield, Belinda Crowe, Angela Van-
Den-Bogerd, Andrew Parsons, Patrick Bourke, Tom Wechsler, Anthony Hooper, lan
Henderson, Chris Aujard and Ron Warmington dated 02.10.14- 15.10.14
255. I POL00202008 Linklaters' Report into Initial Complaint Review and Mediation Scheme Legal Issues
dated 20.03.14
256. I POL00199738 Initial Complaint Review and Mediation Scheme — Draft Settlement Policy dated
December 2013
257. I WBON0001707 Email from Andrew Parsons to Rodric Williams and David Oliver dated 26.03.14
258. I POL00278283 Advice note — Initial Complaint Review and Mediation Scheme — Harm to a retail
business following summary termination dated 26.03.14
259. I POL00043630 Meeting Minutes - Working Group for the Initial Complaint Review and Case Mediation
Scheme — 14.11.14
260. I POL00216273 Letter from JFSA to Sir Anthony Hooper dated 10.11.14
261. I POL00043631 Minute - Working Group for the Initial Complaint Review and Case Mediation Scheme
— 08.12.14
262. I WBONO000915 Email exchange between Andrew Parsons and Tom Wechsler dated 11.11.14-—
12.11.14
263. I WBONO000805 Initial Complaint and Mediation Scheme — Draft Scheme Pack — Part 1: Application
Phase - undated
264. I POL00218712 Initial Complaint Review and Mediation Scheme — Working Group Briefing dated
14.01.15
265. I POL00407979 Initial Complaint and Mediation Scheme — Mediation Briefing dated 14.10.14
266. I POLO0232900 Review letter from CEDR to Patrick Bourke dated 31.07.15
267. I WBON0000408 Email from Andrew Parsons to Simon Richardson dated 03.03.15
268. I WBONO000409 Project Sparrow — Update and Options paper dated March 2015
269. I POL00021908 Email from Andrew Parsons to Belinda Crowe dated 09.02.15
270. I POL00021728 Email exchange between Andrew Parsons and Patrick Bourke dated 05.02.15—
06.02.15
271. I WBONO000921 Email exchange between Andrew Parsons and Patrick Bourke dated 05.02.15
272. I POL00221480 Advice note entitled ‘Termination of Second Sight’ dated 06.02.15
273. I POL00221561 Project Sparrow Sub-Committee — Update and Options paper dated February 2015
274. I POL00407493 Email exchange between Andrew Parsons and Rodric Williams dated 28.06.13
275. I POL00407494 Email chain involving Rodric Williams, Andrew Parsons, Rod Ismay, Simon Baker,
Lesley J Sewell, Andrew Winn and Joanna Jacobson dated 01.07.13
AC_207339548_5 14
WITN10390200
WITN10390200
No. I Document Reference Document
276. I POL00061756 Initial Complaint Review and Mediation Scheme - Post Office Preliminary Investigation
Report dated 02.09.14
277. I POL00021865 Email from Andrew Parsons to Rodric Williams dated 18.08.15
278. I WBONO000948 Email chain involving Mark Underwood, Andrew Parsons and Patrick Bourke dated
12.10.15
279. I POL00006355 A Review on Behalf of the Chairman of Post Office Limited - Concerning the Steps
taken in Response to Various Complaints made by Sub-Postmasters dated 08.02.16
280. I WBONO0000962 Email from Mark Underwood to Andrew Parsons dated 15.04.16
281. I POL00174470 Email chain involving Rodric Williams, Andrew Parsons, Gavin Matthews, Mark
Underwood and John Davitt dated 06.10.15 — 05.05.16
282. I WBONO000952 Email from Mark Underwood to Andrew Parsons dated 26.01.16
283. I WBONO000954 Email exchange between Andrew Parsons and Mark Underwood dated 26.01.16
284. I WBON0000414 Email chain involving Andrew Parsons, Emma Kennedy and Mark Underwood dated
26.01.16
285. I WBONO000955 Email exchange between Andrew Parsons and Mark Underwood dated 26.01.16—
01.02.16
286. I WBONO0000957 Email exchange between Patrick Bourke, Andrew Parsons and Mark Underwood dated
26.01.16 — 01.02.16
287. I WBONO000415 Email exchange between Andrew Parsons, Emma Kennedy and Mark Underwood
dated 26.01.16 — 02.02.16
288. WBONO0000958 Email exchange between Patrick Bourke, Andrew Parsons and Mark Underwood dated
26.01.16 — 10.02.16
289. I POL00239502 Email exchange between Patrick Bourke, Andrew Parsons and Mark Underwood dated
26.01.16 — 17.02.16
290. I WBON0000417 Email chain involving Andrew Parsons, Paul Loraine, Patrick Bourke, and Mark
Underwood dated 26.01.16 — 19.02.16
291. I WBONO0000419 Email chain involving Andrew Parsons, Paul Loraine, Patrick Bourke, and Mark
Underwood dated 26.01.16 — 22.02.16
292. I WBONO000990 Email from Paul Loraine to Mark Underwood dated 05.05.16
293. I POL00241260 Report entitled ‘Investigations conducted by Bond Dickinson LLP, on behalf of Post
Office Ltd, into complaints about the advice provided by Call Handlers at the Network
Business Support Centre (NBSC)' dated 04.05.16
294. I WBON0000420 Email chain involving Katie Watkins, Andrew Parsons, Olivia Moran, Gareth Pole,
Rodric Williams and Elisa Lukas dated 24.02.16 — 25.02.16
295. I POL00028069 Deloitte Board Briefing dated 04.06.16
AC_207339548_5 15
WITN10390200
WITN10390200
No. I Document Reference Document
296. I WBONOO000960 Email exchange between Andrew Parsons and Rodric Williams dated 10.02.16 —
08.03.16
297. I WBONO000965 Email from Rodric Williams to Andrew Parsons dated 15.04.16
298. I WBONO000962 Email form Mark Underwood to Andrew Parsons dated 15.04.16
299. I POL00240675 Engagement Letter dated 09.04.16 Change Order Number 02 11.03.16
300. I WBONO0000339 Email exchange between Rodric Williams and Andrew Parsons dated 15.04.16—
19.04.16
301. I WBONO000984 Email exchange between Rodric Williams and Andrew Parsons dated 26.04.16
302. I WBONO000985 Email Andrew Parsons to Andrew Whitton and Mark Westbrook dated 26.04.16
303. I POL00242882 Letter Bond Dickinson to Deloitte dated 26.04.16
304. I WBONO0000336 Email Andrew Parsons to Rodric Williams and Tony Robinson QC dated 28.06.16
305. I WBON0001674 Group Litigation Order dated 21.03.17
306. I WBONO0001685 CMC Order dated 27.10.17
307. I POLO0117925 CMC Order dated 22.02.18
308. I POL00120352 CMC Order dated 21.06.18
309. I WBONO0001669 CMC Order dated 31.01.19
310. I POL00251998 Decision paper on settlement proposals including comments from Steering Group
members dated 06.12.17
311. I POL00251957 Five Things Document prepared by David Cavender QC dated 14.12.17
312. I POL00252996 Post Office "Black Hat Review" Note prepared by David Cavender QC dated 18
January 2018
313. I POLO0270841 Opinion on the Common Issues dated 10.05.18
314. I WBONO001688 David Cavender QC Speaking Note For POL Board Subcommittee
315. I WBONO000511 Email from Mark Underwood to Jane MacLeod, Angela Van-Den-Bogerd, Mark Davies,
Stuart Nesbit, Patrick Bourke, Rob Houghton, Rodric Williams, Thomas Moran, Tom
Wechsler, Andrew Parsons, Mark Ellis, Melanie Corfield, Nick Beal and Amy Prime
dated 04.12.17
316. I POL00139298 Agenda for Steering Group Meeting on 22.08.16
317. I POL00243195 Agenda for Steering Group Meeting on 14.07.16
318. I POL00139297 Decision Paper for Steering Group Meeting dated 22.08.16
319. I POL00139479 Discussion Paper for Steering Group Meeting on 03.11.17
AC_207339548_5
WITN10390200
WITN10390200
No. I Document Reference Document
320. I POL00251593 Decision Document for Steering Group Meeting on 20.11.17
321. I POL00251596 Decision Paper: Next 12 months for Steering Group Meeting on 20.11.17
322. I POL00261175 Steering Group Paper dated 17.01.2019
323. I POLO0261176 Noting Paper: Appeal process
324. I POL00261172 Noting Paper: Cost of Common Issues Trial for Steering Group Meeting on 17.01.19
325. I POL00259673 Noting Paper: Expert Report of Dr Robert Worden for Steering Group Meeting
28.11.18
326. I POL00024436 Email from Andrew Parsons to Jane MacLeod, Angela Van-Den-Bogerd, Mark Davies,
Stuart Nesbit, Patrick Bourke, Rodric Williams, Thomas Moran, Mark Ellis, Melanie
Corfield, Tom Wechsler and Nick Beal dated 15.12.17
327. I POL00252205 Decision Paper: Proposal for the March 2019 Trial and a Long Term Strategy for the
Group Litigation dated 15.12.17
328. I POLO0024281 Email exchange between Andrew Parsons, Jane MacLeod, Angela Van-Den-Bogerd,
Mark Davies, Stuart Nesbit, Patrick Bourke, Rodric Williams, Thomas Moran, Mark
Ellis, Melanie Corfield, Tom Wechsler and Nick Beal dated 15.12.17
329. I POL00252201 Decision Paper: Proposal for the March 2019 Trial and a Long Term Strategy for the
Group Litigation dated 15.12.17
330. I WBONO000188 Email chain involving Andrew Parsons, Jane MacLeod, Angela Van-Den-Bogerd, Mark
Davies, Stuart Nesbit, Patrick Bourke, Rodric Williams, Thomas Moran, Mark Ellis,
Melanie Corfield, Tom Wechsler and Nick Beal dated 15.12.17 — 17.12.17
331. I WBONO0000328 Email chain involving Andrew Parsons, Jane MacLeod, Angela Van-Den-Bogerd, Mark
Davies, Stuart Nesbit, Patrick Bourke, Rodric Williams, Thomas Moran, Mark Ellis,
Melanie Corfield, Tom Wechsler and Nick Beal dated 15.12.17 -— 17.12.17
332. I WBON0000171 Email chain involving Andrew Parsons, Jane MacLeod, Angela Van-Den-Bogerd, Mark
Davies, Stuart Nesbit, Patrick Bourke, Rodric Williams, Thomas Moran, Mark Ellis,
Melanie Corfield, Tom Wechsler and Nick Beal dated 15.12.17 — 17.12.17
333. I POL00024278 Email chain involving Andrew Parsons, Jane MacLeod, Angela Van-Den-Bogerd, Mark
Davies, Stuart Nesbit, Patrick Bourke, Rodric Williams, Thomas Moran, Mark Ellis,
Melanie Corfield, Tom Wechsler and Nick Beal dated 15.12.17 — 17.12.17
334. I WBONO001341 Email from Andrew Parsons to David Cavender dated 28.10.18
335. I POL00024633 Email exchange between Jane MacLeod, Andrew Parsons and Thomas Moran dated
19.09.17
336. I POLOO117761 Draft Litigation Options Paper dated 19.09.17
337. I POL00024235 Email chain involving Andrew Parsons, Angela Van Den Bogerd, Jane MacLeod and
Rodric Williams dated 24.07.18-25.07.18
338. I POL00358137 Draft Mitigation Actions Paper dated 25.07.18
AC_207339548_5 7
WITN10390200
WITN10390200
No. I Document Reference Document
339. I WBONO001658 Email chain involving Andrew Parsons, Emma Deas, Rodric Williams and Catherine
Emanuel dated 11.11.19
340. I POL00288584 Draft Group Litigation Update Paper dated 11.11.19 (for board meeting on 13.11.19)
341. I WBONO001663 Email chain involving Andrew Parsons, Rodric Williams, Patrick Bourke, Angela Van
Den Bogerd, Mark Underwood, Ben Foat and Alan Watts dated 04.12.19 to 09.12.19
342. I POLO0289960 Draft Group Litigation Update Paper dated 09.12.19 (for board meeting on 10.12.19)
343. I WBONO000510 Email exchange between Andrew Parsons and Anthony de Garr Robinson dated
01.11.17 — 07.11.17
344. I POLO0249671 Email from Victoria Brooks to Rodric Williams dated 07.07.17
345. I POL00249674 Draft Postmaster Litigation Paper dated 07.07.17 (for Group Executive meeting on
13.07.17)
346. I POL00006384 Email exchange between Amy Prime, Andrew Parsons and Anthony de Garr Robinson
dated 28.09.17
347. I POL00006499 Email from Andrew Parsons to Jane MacLeod dated 28.09.17
348. I POL00006380 Paper for 11.09.17 Steering Group meeting
349. I POL00250513 Litigation Strategy Options document dated 12.09.17
350. I POL00250466 Decision Paper: Steering Group Meeting dated 11.09.17
351. I POL00006503 Litigation Strategy Options: Steering Group Meeting dated 11.09.17
352 POL00252386 Letter Womble Bond Dickinson to Freeths dated 18.12.17
353. I WBON0001377 Letter Freeths to Womble Bond Dickinson dated 10.01.18
354. I WBON0001378 Letter Womble Bond Dickinson to Freeths dated 17.01.18
355. I POLO0006360 Decision Paper: Does Post Office engage in further mediation? Dated 08.07.16
356. I POL00247209 Decision Paper: Steering Group Meeting dated 14.02.17
357. I POL00139476 Noting Paper : Steering Group Meeting: 3 November 2017
358. I WBON0001230 CMC Order dated 08.02.18
359. I POLO00006382 Speaking Note For Sub-Committee Meeting 15.05.18
360. I POLO0259669 Decision Paper: Steering Group dated 28.11.18
361. I POL00265780 Letter from Womble Bond Dickinson to Freeths dated 07.12.18
362. I POL00260751 Letter from Freeths to Womble Bond Dickinson dated 21.12.18
363. I POL00265783 Letter from Womble Bond Dickinson to Freeths dated 09.01.19
AC_207339548_5
WITN10390200
WITN10390200
No. I Document Reference Document
364. I POL00262338 Letter from Freeths to Womble Bond Dickinson dated 17.01.19
365. I POL00275113 Post Office Group Litigation Settlement Briefing dated 19.05.19
366. I POL00023690 Email from Amy Prime to Mark Underwood dated 14.06.19
367. I POL00000444 Alan Bates & Others And Post Office Limited: Fourth Witness Statement of Andrew
Paul Parsons dated 09.10.17
368. I POL00041510 Email from Freeths to Bond Dickinson dated 16.10.17
369. I POL00041509 Email from Andrew Parsons to Rodric Williams dated 16.10.17
370. I WBON0001194 Letter from Bond Dickinson to Freeths dated 01.09.17
371. I WBON0001216 Letter from Freeths to Bond Dickinson dated 13.10.17
372. I WBONO001215 Email chain involving James Hartley, Peter O'Connell, Andrew Parsons, Elisa Lukas,
Amy Prime, Anthony de Garr Robinson and Owain Draper
373. I WBONO0000191 Email Chain involving Imogen Randall, PETER O'Connell, Andrew Parsons, Owain
Draper and Amy Prime dated 16.10.17
374. I WBONO001217 Email Chain involving Ann Harries, clerks@hendersonchambers.co.uk, Andrew
Parsons, Rob Smith, Owain Draper, Anthony de Garr Robinson dated 08.11.17
375. POL00041527 Email from Rodric Williams to Jane MacLeod, Melanie Corfield, Mark Underwood,
Mark Davies and Thomas Moran dated 09.11.17
376. I WBONO000329 Email Andrew Parsons to James Hartley dated 12.12.18
377. POL00041136 Email exchange between Andrew Parsons, Rodric Williams and Mark Underwood
dated 20.04.16 — 21.04.16
378. I WBONO000987 Email Chain involving Rodric Williams, Alwen Lyons, Craig Tuthill, Lin Norbury, John
Breeden, Joe Connor, hector Campbell, Angela Van-Den-Bogerd, Kathryn Alexander,
Shirley Hailstones, Chris Broe, Andy gardner, Julie George, John M Scott, Nick Beal,
Anne Allaker and Andrew Parsons dated 20.04.16 — 03.05.16
379. I WBONO000988 Disclosure of Documents in Litigation dated -12.14
380. I POL00241034 Alan Bates and Others and Post Office Limited Schedule of Claimants
381. I WBONO000151 Email From Tom Porter to Rodric Williams dated 09.08.16
382. I WBON0001002 Email Chain involving Rodric Williams , Alwen Lyons, Craig Tuthill, Lin Norbury, John
Breeden, Joe Connor, Hector Campbell, Angela Van-Den-Bogerd, Kathryn Alexander,
Shirley Hailstones, Chris Broe, Andy Garner, Julie George, John Scott, Nick Bea, Anne
Allaker, Tom Porter and Dave King dated 20.04.16 - 31.05.16
383. I WBONO001015 Email Chain involving Rodric Williams , Alwen Lyons, Craig Tuthill, Lin Norbury, John
Breeden, Joe Connor, Hector Campbell, Angela Van-Den-Bogerd, Kathryn Alexander,
Shirley Hailstones, Chris Broe, Andy Garner, Julie George, John Scott, Nick Bea, Anne
Allaker, Tom Porter and Dave King dated 20.04.16 - 03.06.16
AC_207339548_5 19
WITN10390200
WITN10390200
No. I Document Reference Document
384. I POL00139309 Postmaster Litigation Steering Group Actions — undated
385. I WBON0000982 Email from Andrew Parsons to Rodric Williams dated 26.04.16
386. I WBONO0000981 Letter from Bond Dickinson to Fujitsu dated 26.04.16
387. I WBONOO00992 Email from Tom Porter to Rodric Williams dated 06.05.16
388. I WBONO000154 Email exchange between Elisa Lukas, Rodric Williams and Andrew Parsons dated
15.11.16 — 16.11.16
389. I POL00041378 Email exchange between Rodric Williams and Amy Prime dated 28.11.16
390. I POL00006436 Decision Paper: Steering Group Meeting dated 08.08.17
391. I POL00139383 Decision Paper: Steering Group Meeting dated 24.05.17
392. I WBONO0001686 Decision Paper: Steering Group Meeting dated 04.01.17
393. I POLO0006405 Decision Paper: Steering Group Meeting dated 24.05.17
394. I POL00006470 Decision Paper: Steering Group Meeting dated 11.09.17
395. I POL00269447 Decision Paper: POLSAP Data Hosted by Fujitsu dated 22.03.19
396. I POL00278526 Decision Paper: Deletion of Data Held on Brands Database dated 08.08.19
397. I POL00139652 Decision Paper: Preservation of Data Stored on Post Office's File Servers dated
17.07.19
398. I POL00139650 Decision Paper: Preservation of POLSAP dated 17.07.19
399. I POL00288913 Decision Paper: Deletion of data held on Fujitsu Telecoms dated 19.11.19
400. I WBONO001013 Email Chain involving Kerry Moodie, John Scott, Amy Quirk, Helen Dickinson and
Simon Hutchinson dated 19.05.16 — 01.06.16
401. I WBONO001001 Draft Letter from Post Office Information Rights Team to Katherine McAlerney dated
27.05.16
402. I POL00038852 Email from Amy Prime to Rodric Williams dated 10.05.16
403. I POL00241140 Letter from Freeths to Rodric Williams dated 28.04.16
404, POL00110507 Letter from Bond Dickinson to Freeths dated 28.07.16
405. I WBONO000466 Conduct of Criminal Investigations Policy
406. I WBONO000443 Email exchange between Andrew Parsons and Brian Altman dated 18.07.16 —
22.07.16
407. I WBONO0000464 Email chain involving Amy Prime, Amy Quirk, Jane MacLeod and Andrew Parsons
dated 19.09.16 — 29.09.16
408. I WBONO0000465 Email from Amy Prime to Andrew Parsons dated 05.10.16
AC_207339548_5 20
WITN10390200
WITN10390200
No. I Document Reference Document
409. I WBONO0000467 Email exchange between Amy Prime and Andrew Parsons dated 05.10.16
410. I POL00022636 Email from Andrew Parsons to Rodric Williams and Jane MacLeod dated 18.05.16
411. I POL00156685 Briefing Note for Counsel
412. I WBONO000993 Email exchange between Andrew Parsons, Rodric Williams and Paul Loraine dated
07.05.16 — 11.05.16
413. I WBONO000179 Email from Paul Loraine to Gavin Matthews dated 10.05.16
414. I WBONO0000157 Email from Tom Porter to Andrew Parsons dated 07.06.16
415. I POL00006601 Letter from Bond Dickinson to Post Office Limited dated 21.06.16
416. I POL00242402 Email exchange between Andrew Parsons and Anthony de Garr Robinson dated
08.06.16
417. I POL00242578 Draft Letter from Bond Dickinson to Post Office Limited dated 16.06.16
418. I POL00041242 Email exchange between Gavin Matthews and Rodric Williams dated 16.06.16 —
17.06.16
419. I WBONOOO0995 Email from Andrew Parsons to Anthony de Garr Robinson dated 24.05.16
420. I POL00140216 Email exchange between Andrew Parsons and Anthony de Garr Robinson dated
01.06.16
421. I POL00041770 Email chain involving Patrick Bourke, Elizabeth O'Neill, Rodric Williams, Helen
Lambert, Jane MacLeod and Andrew Parsons dated 23.02.18— 11.05.18
422 WBONO0000160 Email exchange between Andrew Parsons and Tom Cooper dated 30.04.19 to
02.05.19
423. I WBONOO000648 Email exchange between Rodric Williams, Andrew Parsons and Jonathan Gribben
dated 07.01.19
424. I WBONO000662 Email chain involving Jane MacLeod, Patrick Bourke, Mark Underwood, Amy Prime,
Tom Beezer, Andrew Parsons dated 14.03.19
425. I POL00023809 Email chain involving Jane MacLeod, Patrick Bourke, Mark Underwood, Amy Prime
and Ben Beabey dated 14.03.19- 15.03.19
426. I POL00024241 Email exchange between Jane MacLeod, Andrew Parsons and Rodric Williams dated
08.06.18 — 11.06.18
427. I POL00041825 Email chain involving Rodric Williams, Elizabeth O'Neill, Helen Lambert, Jane
MacLeod, Mark Underwood and Andrew Parsons dated 30.05.18
428. I WBON0001306 Draft Email to UKGI from Rodric Williams dated 23.05.18
429. I WBONOO000705 Email exchange between Andrew Parsons, Jane MacLeod and Rodric Wiliams dated
10.04.19 — 11.04.19
AC_207339548_5 2
WITN10390200
WITN10390200
No. I Document Reference Document
430. I WBON0001249 Email chain involving Andrew Parsons, James Hartley, Imogen Randall, Rodric
Williams, Mark Underwood and Amy Prime dated 08.06.18 — 11.06.18
431. I WBONO001417 Email chain involving Amy Prime, Mark Underwood, Rodric Williams, Angela Van-Den-
Bogerd and Andrew Parsons dated 12.02.19
432. I WBONO000691 Email chain involving Richard Watson, Jane MacLeod, Rodric Williams and Andrew
Parsons dated 25.03.19
433. I POL00023301 Email chain involving Richard Watson, Jane MacLeod, Rodric Williams and Andrew
Parsons dated 25.03.19 — 29.03.19
434. I WBONO0001643 Email chain involving Ben Foat, Rodric Williams, Angela Van-Den-Bogerd, Andrew
Parsons, Tom Beezer, Patrick Bourke and Mark Underwood dated 24.06.19
435. I WBONO000641 Email chain involving Charlie Temperley, Andrew Parsons, Dave Panaech and Amy
Prime dated 07.1218 — 09.12.19
436. I WBONO0000647 Email from Andrew Parsons to Charlie Temperley dated 07.01.19
437. I WBONOO00719 Email chain involving Angelique Richardson, Angela Fraser, Andrew Parsons, Rodric
Williams, Angela Van-Den-Bogerd, Mark Underwood and Catherine Emmanuel dated
30.09.19
438. I WBONO001248 Email exchange between Andrew Parsons, Rodric Williams, Mark Underwood, Jane
MacLeod, Thomas Moran and Melanie Corfield dated 31.05.18 — 01.06.18
439. I WBONO001068 Email chain involving Laura Thompson, Patrick Bourke, Mark Underwood, Andrew
Parsons and Rodric Williams dated 03.01.17 — 09.01.17
440. I WBONO001179 Email exchange between Melanie Corfield, Andrew Parsons and Mark Underwood
dated 17.07.17
441. I POL00041684 Email chain involving Patrick Bourke, Elizabeth O'Neill, Jane MacLeod, Rodric
Williams and Andrew Parsons dated 23.02.18 — 21.03.18
442. I WBONO000524 Email chain involving Patrick Bourke, Elizabeth O'Neill, Jane MacLeod, Rodric
Williams, Andrew Parsons and Mark Underwood dated 23.02.18 — 21.03.18
443. I POLO0041687 Email chain involving Patrick Bourke, Elizabeth O'Neill, Jane MacLeod, Rodric
Williams and Andrew Parsons dated 23.02.18 — 22.03.18
444. I WBONO000525 Email chain involving Patrick Bourke, Elizabeth O'Neill, Jane MacLeod, Rodric
Williams, Andrew Parsons, Amy Prime, Paul Stewart and Jonathan Gribben dated
23.02.18 — 22.03.18
445. I WBONO0000528 Email from Amy Prime to Andrew Parsons dated 23.03.18
446. I POL00041695 Email from Rodric Williams to Jane MacLeod, Patrick Bourke, Mark Underwood,
Thomas Moran, Veronica Branton, Andrew Parsons, Amy Prime and Ben Foat dated
24.03.18
447. I POLO0254174 Post Office Group Litigation Information Sharing Protocol
AC_207339548_5 22
WITN10390200
WITN10390200
No. I Document Reference Document
448. POL00041697 Email chain involving Rodric Williams, Jane MacLeod, Patrick Bourke, Mark
Underwood, Thomas Mora, Veronica Branton, Andrew Parsons, Amy Prime and Ben
Foat dated 24.03.18
449. I WBONO0001236 Email chain involving Patrick Bourke, Elizabeth O'Neill, Rodric Williams, Helen
Lambert, Andrew Parsons and Mark Underwood dated 23.02.18 — 20.04.18
450. I POL00041760 Email chain involving Patrick Bourke, Elizabeth O'Neill, Rodric Williams, Helen
Lambert, Jane MacLeod, Andrew Parsons, and Mark Underwood dated 23.02.18—
20.04.18
451. I WBONO001240 Email chain involving Patrick Bourke, Elizabeth O'Neill, Rodric Williams, Helen
Lambert, Jane MacLeod and Andrew Parsons dated 23.02.18— 24.04.18
452. I WBONO001241 Email from Andrew Parsons to Jane Macleod and Rodric Williams dated 26.04.18
453. I POLO0041772 Email chain involving Patrick Bourke, Elizabeth O'Neill, Rodric Williams, Helen
Lambert, Jane MacLeod, and Andrew Parsons dated 23.02.18— 11.05.18
454. I WBONO001244 Email from Andrew Parsons to Rodric Williams dated 14.05.18
455. I WBONO0001245 Email from Rodric Williams to Andrew Parsons dated 17.05.18
456. I WBONO0001251 Email chain involving Rodric Williams, Helen Lambert, Elizabeth O'Neill, Jane
MacLeod, Andrew Parsons, Mark Underwood and Patrick Bourke dated 11.06.18
457. I WBONO001648 UKGI / Post Office Limited Information Sharing Protocol
458. I POL00242335 Draft Workplan — Actions and Timings (For Letter of Claim Response)
459. I POL00243124 Post Office Group Action — Bond Dickinson Workplan (For Letter of Claim Response)
460 WBONO0001019 Email exchange between Andrew Parsons and Rodric Williams dated 27.06.16
461. I WBONO0000431 Email exchange between Andrew Parsons and Paul Loraine dated 10.07.16—
14.07.16
462 WBON0000427 Email chain involving Tom Porter, Amy Prime and Andrew Parsons dated 24.06.18 —
08.07.16
463. I WBONO0000424 Email from Andrew Parsons to Paul Loraine dated 13.06.16
464. I POL00243114 Decision 1: Does Post Office Address in Detail the "Bugs" in Horizon Identified by
Second Sight
465. I WBONO001025 Email from Andrew Parsons to Anthony de Garr Robinson dated 10.07.16
466. I POL00025373 Email from Andrew Parsons to Anthony de Garr Robinson dated 08.07.16
467. I WBONO000426 Email exchange between Andrew Parsons and Anthony de Garr Robinson dated
27.06.16
468. I WBONO001021 Email chain involving Imogen Randall, Andrew Parsons and Anthony de Garr
Robinson dated 04.07.16 — 05.07.16
AC_207339548_5 23
WITN10390200
WITN10390200
No. I Document Reference Document
469. I WBONO0001022 Post Office: Thoughts on Draft Letter of Response — Note Prepared by Anthony de
Garr Robinson QC dated 5 July 2016
470. I WBONO001024 Email exchange between Andrew Parsons Anthony de Garr Robinson and Owain
Draper dated 06.07.16
471. I WBONO0000432 Email chain involving Andrew Parsons, Anthony de Garr Robinson, Tiffany Redhead,
Paul Loraine and Amy Prime dated 13.07.16— 15.07.16
472. I WBONO001031 Email chain involving Andrew Parsons, Anthony de Garr Robinson, Tiffany Redhead,
and Paul Loraine dated 13.07.16 — 16.07.16
473. I WBONO0000434 Email chain involving Andrew Parsons, Anthony de Garr Robinson, Tiffany Redhead,
and Paul Loraine dated 13.07.16 — 16.07.16
474. I WBONO001023 Email exchange between Paul Loraine, Brian Altman and Andrew Parsons dated
08.06.16 — 06.07.16
475. I WBON0001033 Email from Andrew Parsons to Brian Altman dated 18.07.16
476. I WBONO001047 Email exchange between Andrew Parsons and Brian Altman dated 18.07.16 —
22.07.16
477. I WBONO000423 Email from Andrew Parsons to Jonathan Gribben dated 12.06.16
478. I POL00041238 Email exchange between Andrew Parsons, Patrick Bourke, Rodric Williams and Mark
Underwood dated 14.06.16 — 17.06.16
479. I WBONO0001018 Email chain involving Paul Loraine, Kathryn Alexander, Shirley Hailstones, Andrew
Winn, Dave King and Mark Underwood dated 17.06.16 - 24.06.16
480. I POL00243355 Postmaster Litigation Steering Group Actions
481. I WBON0000435 Email from Andrew Parsons to Thomas Moran, Rodric Williams, Angela Van-Den-
Bogerd, Patrick Bourke, Mark Davies, Rob Houghton, Tom Wechsler, Nick Sambridge,
Jane MacLeod and Mark Underwood dated 16.07.16
482. I WBONO0001038 Email chain involving Andrew Parsons, Kathryn Alexander, Shirley Hailstones and
Angela Van-Den-Bogerd dated 18.07.16 — 19.07.16
483. I WBONO001036 Email exchange between Jessica Madron and Andrew Parsons dated 16.07.16—
19.07.16
484 WBON0001048 Email exchange between Jessica Madron, Andrew Parsons and Rodric Williams dated
16.07.16 — 25.07.16
485. I WBONO001054 Email exchange between Jessica Madron, Andrew Parsons and Rodric Williams dated
16.07.16 — 25.07.16
486. I WBON0000436 Steering Group Spreadsheet Relating To Reviews Of Letter of Response
487. I WBON0001050 Email from Andrew Parsons to Rodric Williams, Jane MacLeod and Thomas Moran
dated 25.07.16
AC_207339548_5 24
WITN10390200
WITN10390200
No. I Document Reference Document
488. I POL00041259 Email from Andrew Parsons to Jane MacLeod, Mark Underwood, Angela Van-Den-
Bogerd, Rob Houghton, Patrick Bourke, Tom Wechsler, Mark Davies, Melanie Corfield,
Thomas Moran and Rodric Williams dated 27.07.16
489. I WBONO0001061 Email from Jane MacLeod to Andrew Parsons dated 28.07.16
490. I POL00408673 Email chain involving Andrew Parsons, Brian Altman and Gavin Matthews dated
18.07.16 — 25.07.16
491. I WBONO0000444 Email chain involving Andrew Parsons, Brian Altman and Gavin Matthews dated
18.07.16 — 26.07.16
492. I WBONO000445 Email chain involving Andrew Parsons, Brian Altman and Gavin Matthews dated
18.07.16 — 26.07.16
493. I WBONO000446 Review of Post Office Limited Criminal Prosecutions Conducted by Brian Altman QC
dated 26 July 2016
494. I POL00022754 Email from Andrew Parsons to Rodric Williams, Jane MacLeod, Patrick Bourke and
Mark Underwood dated 26.07.16
495. I POL00112884 Review of Post Office Limited Criminal Prosecutions
496. I WBON0000470 Email chain involving Andrew Parsons, Paul Loraine, Brian Altman, Gavin Matthews
and Amy Prime dated 18.07.16 — 26.07.16 and 25.10.16
497. I WBONO0000450 Email chain involving Andrew Parsons, Brian Altman, Gavin Matthews and Amy Prime
dated 18.07.16 — 26.07.16
498. I WBONO001016 Email from Andrew Parsons to Patrick Bourke and Rodric Williams dated 14.06.16
499. I WBONO0001644 Email chain involving Mark Westbrook, Rodric Williams, Patrick Bourke, Mark
Underwood, Andrew Parsons, Paul Loraine and Jonathan Gribben dated 08.07.16
500. I POL00243100 Sparrow Interim Report dated 08.07.16
501. I WBON0000430 Email chain involving Andrew Parsons, Jane MacLeod, Rodric Williams, Patrick
Bourke and Jonathan Gribben dated 13.07.16— 14.07.16
502. I WBONO001030 Email from Andrew Parsons to Jane MacLeod, Rodric Williams and Patrick Bourke
dated 13.07.16
503. I WBON0001041 Rider: Remote Access
504. I WBONO001040 Email from Andrew Parsons to Anthony de Garr Robinson dated 20.07.16
505. I WBONO0000438 Email exchange between Andrew Parsons and Anthony de Garr Robinson dated
20.07.16
506. I WBONO001042 Email exchange between Andrew Parsons and Anthony de Garr Robinson dated
21.07.16
507. I WBON0001044 Rider: Remote Access
508. I POL00029997 Rider: Remote Access
AC_207339548_5 25
WITN10390200
WITN10390200
No. I Document Reference Document
509. I POL00408665 Email exchange between Andrew Parsons and Anthony de Garr Robinson dated
20.07.16 — 21.07.16
510. I POL00243366 Rider: Remote Access
511. I WBON0000439 Email exchange between Andrew Parsons and Anthony de Garr Robinson dated
20.07.16 — 21.07.16
512. I POLO0024801 Email from Andrew Parsons to Thomas Moran, Rodric Williams, Angela Van-Den-
Bogerd, Patrick Bourke, Mark Davies, Rob Houghton, Tom Wechsler, Nick Sambridge,
Jane MacLeod and Mark Underwood dated 21.07.16
513. I WBONO0000441 Email chain involving Andrew Parsons, Thomas Moran, Rodric Williams, Angela Van-
Den-Bogerd, Patrick Bourke, Mark Davies, Rob Houghton, Tom Wechsler, Nick
Sambridge, Jane MacLeod and Mark Underwood dated 21.07.16
514. I POL00024876 Email chain involving Andrew Parsons, Thomas Moran, Rodric Williams, Angela Van-
Den-Bogerd, Patrick Bourke, Mark Davies, Rob Houghton, Tom Wechsler, Nick
Sambridge, Jane MacLeod and Mark Underwood dated 21.07.16
515. I WBON0000440 Email chain involving Andrew Parsons, Thomas Moran, Rodric Williams, Angela Van-
Den-Bogerd, Patrick Bourke, Mark Davies, Rob Houghton, Tom Wechsler, Nick
Sambridge, Jane MacLeod and Mark Underwood dated 21.07.16
516. POL00024876 Email chain involving Andrew Parsons, Thomas Moran, Rodric Williams, Angela Van-
Den-Bogerd, Patrick Bourke, Mark Davies, Rob Houghton, Tom Wechsler, Nick
Sambridge, Jane MacLeod and Mark Underwood dated 21.07.16
517. I WBON0001045 Email exchange between Michael Harvey, Rodric Williams and Andrew Parsons dated
15.04.16 — 21.07.16
518. I POL00408671 Email from Andrew Parsons to Mark Westbrook and Lewis Keating dated 21.07.19
519. I POL00243580 Email exchange between Andrew Parsons to Mark Westbrook and Lewis Keating
dated 21.07.19 - 22.07.19
520. I WBON0000442 Email exchange between Michael Harvey, Rodric Williams and Andrew Parsons dated
15.04.16 — 22.07.16
521. I WBONO000447 Email exchange between Michael Harvey, Rodric Williams and Andrew Parsons dated
15.04.16 — 26.07.16
522. I WBON0000449 Email exchange between Michael Harvey, Rodric Williams and Andrew Parsons dated
15.04.16 — 27.07.16
523. I POL00023428 Email exchange between Michael Harvey, Rodric Williams, Andrew Parsons, Jane
MacLeod and Rob Houghton dated 15.04.16- 27.07.16
524. I POL00025320 Email chain involving James Davidson, Rodric Williams, Mark Underwood, Patrick
Bourke, Jane MacLeod and Andrew Parsons dated 14.04.14- 26.07.16
525. I WBON0000448 Email chain involving James Davidson, Rodric Williams, Mark Underwood, Patrick
Bourke, Jane MacLeod, Andrew Parsons and Rob Houghton dated 14.04.14-
26.07.16
AC_207339548_5 26
WITN10390200
WITN10390200
No. I Document Reference Document
526. POL00024824 Email chain involving James Davidson, Rodric Williams, Mark Underwood, Patrick
Bourke, Jane MacLeod, Andrew Parsons, Rob Houghton and Angela Van-Den-Bogerd
dated 14.04.14 — 26.07.16
527. I POL00024794 Email chain involving James Davidson, Rodric Williams, Mark Underwood, Patrick
Bourke, Jane MacLeod, Andrew Parsons, and Rob Houghton dated 14.04.14-
26.07.16
528. I POL00024828 Email chain involving James Davidson, Rodric Williams, Mark Underwood, Patrick
Bourke, Jane MacLeod, Andrew Parsons, Rob Houghton, Angela Van-Den-Bogerd
and Thomas Moran dated 14.04.14 — 26.07.16
529. I POL00357378 Rider: Remote Access
530. I WBON0000452 Email chain involving Amy Prime, Mark underwood, Angela Van-Den-Bogerd, Rob
Houghton, Jane MacLeod, Rodric Williams, Patrick Bourke, Tom Wechsler, Mark
Davies, Melanie Corfield, Thomas Moran dated 27.07.16
531. I WBON0000453 Email chain involving Amy Prime, Mark underwood, Angela Van-Den-Bogerd, Rob
Houghton, Jane MacLeod, Rodric Williams, Patrick Bourke, Tom Wechsler, Mark
Davies, Melanie Corfield, Thomas Moran dated 27.07.16
532. I WBONO000454 Email chain involving Amy Prime, Mark underwood, Angela Van-Den-Bogerd, Rob
Houghton, Jane MacLeod, Rodric Williams, Patrick Bourke, Tom Wechsler, Mark
Davies, Melanie Corfield, Thomas Moran dated 27.07.16
533. I WBONO000455 Email chain involving Amy Prime, Mark underwood, Angela Van-Den-Bogerd, Rob
Houghton, Jane MacLeod, Rodric Williams, Patrick Bourke, Tom Wechsler, Mark
Davies, Melanie Corfield, Thomas Moran dated 27.07.16
534. I WBONO000456 Email chain involving Amy Prime, Mark Underwood, Angela Van-Den-Bogerd, Rob
Houghton, Jane MacLeod, Rodric Williams, Patrick Bourke, Tom Wechsler, Mark
Davies, Melanie Corfield, Thomas Moran dated 27.07.16
535. I WBONO001057 Email chain involving Amy Prime, Mark underwood, Angela Van-Den-Bogerd, Rob
Houghton, Jane MacLeod, Rodric Williams, Patrick Bourke, Tom Wechsler, Mark
Davies, Melanie Corfield, Thomas Moran dated 27.07.16
536. I POLO0408686 Email chain involving Rodric Williams, Micheal Harvey, Andrew Parsons and Amy
Prime dated 15.04.16 — 27.07.16
537. I WBON0001672 Rider: Remote Access
538. I WBONO0000457 Email exchange Andrew Parsons, Mark Westbrook and Lewis Keating dated 19.07.19
— 28.07.16
539. I WBONO000458 Email exchange Andrew Parsons, Mark Westbrook and Lewis Keating dated 19.07.19
— 29.07.16
540. I POL00408699 Email exchange Andrew Parsons, Mark Westbrook and Lewis Keating dated 19.07.19
— 29.07.16
541. I WBONO001055 Email chain involving Rodric Williams, Micheal Harvey and Andrew Parsons dated
15.04.16 - 27.07.16
542. I POL00408688 Email from Andrew Parsons to Anthony de Garr Robinson dated 27.07.16
AC_207339548_5 27
WITN10390200
WITN10390200
No. I Document Reference Document
543. I POL00022663 Email from Jane MacLeod to Paula Vennells and Alisdair Cameron dated 28.07.16
544. I POL00022664 Email chain involving Rob Houghton, Alisdair Cameron and Jane MacLeod dated
27.07.16
545. I POLO0250455 Amended Generic Particulars of Claim
546. I WBONO001061 Email from Owen Draper to Andrew Parsons dated 19.05.17
547. I WBON0001081 Outline Defence dated 19.05.17
548. I WBONO001071 Email from Amy Prime to Anthony De Garr Robinson and Owain Draper dated
24.03.17
549. I WBON0001677 Decision Paper on case strategy for 14.02.17 Steering Group meeting
550. I WBON0000474 Email from Amy Prime to Andrew Parsons, Jonathan Gribben and Elisa Lukas dated
14.03.17
551. I WBONO000478 Email From Amy Prime to Kathryn Alexander and Shirley Hailstones dated 29.03.17
552. I POL00023448 Email exchange involving Mark Underwood, Andrew Parsons Jonathon Gribben,
Alisdair Cameron and Amanda Radford dated 04.03.17
553. I POL00023449 Suspense Accounts Questions
554. WBON0001079 Email exchange involving Mark Underwood, Mark Westbrook, Lewis Keating, Andrew
Parsons, Jonathan Gribben and Alisdair Cameron dated 16.05.17
555. I WBONO0001080 Suspense Accounts Briefing Note
556. WBON0001085 Email From Amy Prime to Anthony de Garr Robinson and Owain Draper dated
30.05.17
557. I POL00249406 Email from Amy Prime to Anthony de Garr Robinson and Owain Draper dated
05.06.17
558. I WBONO001116 Email From Amy Prime to Anthony de Garr Robinson and Owain Draper dated
26.06.17
559. I WBONO0000481 Email exchange involving Amy Prime, Shirley Hailstones and Kathryn Alexander dated
02.06.17 to 05.06.17
560. I WBONOO001112 Email exchange involving Rodric Williams, Elisa Lukas and Chris Jay dated 14.06.17
561. I WBONO001115 Email exchange involving Andrew Parsons, Rodric Williams, Elisa Lukas, Amy Prime
and Pete Newsome dated 20.06.17 to 21.06.17
562. I WBON0000484 Email From Amy prime to Andrew Parsons, and Kizzie Fenner dated 30.06.17
563. I WBONO001126 Email from Andrew Parsons to Anthony De Garr Robinson and Owain Draper dated
03.07.17
564. I WBON0001083 Email from Andrew Parsons to Rodric Williams dated 23.05.17
AC_207339548_5 28
WITN10390200
WITN10390200
No. I Document Reference Document
565. I WBONO001121 Email from Andrew Parsons to Rodric Williams and Mark Underwood dated 28.06.17
566. I WBONO0000485 Email from Anthony de Garr Robinson to Andrew Parsons and Owain Draper dated
04.07.17
567. I WBON0000487 Email exchange involving Andrew Parsons, Amy Prime, Elisa Lukas and Anthony De
Garr Robinson dated 04.07.17
568. I WBONO0000489 Email exchange involving Andrew Parsons, Amy Prime, Elisa Lukas and Anthony De
Garr Robinson dated 04.07.17
569. I WBON0001145 Email from Rodric Williams to Andrew Parsons dated 07.07.17
570. I POLO00249670 Draft Generic Defence— dated 07.07.17
571. I WBON0000491 Email from Andrew Parsons to Christopher Jay dated 04.07.17
572. I WBON0000492 Draft Defence — Horizon Related Sections — dated 04.07.17
573. I WBONO001128 Email exchange involving Andrew Parsons, Mark Westbrook and Christopher Jay
dated 04.07.17
574. I WBONO001147 Email exchange between Andrew Parsons, Rodric Williams, Amy Prime and
Christopher Jay dated 11.07.24
575. I WBON0001130 Email from Amy Prime to Kathryn Alexander, Shirley Hailstones dated 05.07.17
576. I WBON0001132 Draft Generic Defence — Rider for Kathryn Alexander and Shirley Hailstones dated
05.07.17
577. I WBON0001138 Email from Amy Prime to Gayle Peacock and Kendra Dickinson dated 06.07.17
578. I WBONO001140 Draft Generic Defence —- Rider for Gayle Peacock and Kendra Dickinson dated
06.07.17
579. I WBON0000493 Email Amy Prime to Angela Van Den Bogerd dated 10.07.17
580. I WBON0000492 Draft Defence — Horizon Related Sections — dated 11.07.17
581. I WBONO000496 Email exchange involving Anthony de Garr Robinson, Amy Prime and Owain Draper
dated 10.07.17
582. I WBONO000497 Email chain between Andrew Parsons, Amy Prime and Anthony de Garr Robinson
dated 11.07.17
583. I WBON0001683 Email chain between Andrew Parsons, Amy Prime and Anthony de Garr Robinson
dated 11.07.17 to 12.07.17
584. I WBONO001157 Draft Defence dated 12.07.17
585. I WBON0001158 Email exchange involving Mark Underwood Jane MacLeod, Angela , Van Den Bogerd,
Mark R Davies, Andrew Parsons, Mark Ellis, Melanie Corfield, Thomas Moran, Stuart
Nesbit, Patrick Bourke, Rob Houghton, Rodric Williams and Tom Wechsler dated
11.07.17 - 12.07.17
AC_207339548_5 29
WITN10390200
WITN10390200
No. I Document Reference Document
586. I WBONO0000498 Email from Amy Prime to Andrew Parsons dated 12.07.17
587. I WBONO000500 Email exchange involving Mark Underwood Jane MacLeod, Angela Van Den Bogerd,
Mark R Davies, Andrew Parsons, Mark Ellis, Melanie Corfield, Thomas Moran, Stuart
Nesbit, Patrick Bourke, Rob Houghton, Rodric Williams and Tom Wechsler dated
11.07.17 — 13.07.17
588. I WBON0001150 ae chain between Amy Prime and Angel Van Den Bogerd dated 10.07.17—
589. I WBONO001153 Draft Generic Defence — with comments from Angela Van Den Bogerd dated 11.11.17
590. I WBON0001154 Email From Mark Westbrook to Andrew Parsons dated 12.07.17
591. I POLO0110670 Draft Defence Horizon Related Sections dated 04.07.17
592. I WBONO0000499 Draft Generic Defence including comments from Deloitte, Fujitsu, Amy Prime and
Angela Van Den Bogerd dated 12.07.17
593. I WBONO001147 anal exchange involving Amy Prime, Christopher Jay and Andrew Parsons dated
594. I WBONO001161 Email exchange involving Amy Prime, Christopher Jay and Andrew Parsons dated
11.07.17 — 12.07.17
595. I POL00249903 Draft Generic Defence with Fujitsu amendments dated 12.07.17
596. I POL00249919 Email from Andrew Parsons to Anthony de Garr Robinson dated 13.07.17
597. I WBONO001163 Email from Mark Underwood to Andrew Parsons and Jonathan Gribben 13.07.17
598. I WBON0001164 Email from Andrew Parsons to Anthony de Garr Robinson dated 13.07.17
599. I WBONO001165 Suspense Account rider — dated 13.07.17
600. I POL00024627 Email exchange involving Jane MacLeod, Andrew Parsons, Mark Underwood Angela
Van Den Bogerd, Mark R Davies, Mark Ellis, Melanie Corfield, Thomas Moran, Stuart
Nesbitt, Patrick Bourke, Rob Houghton and Rodric Williams dated 11.07.17- 14.07.17
601. I WBON0000501 Email from Owain Draper to Andrew Parsons dated 14.07.17
602. I WBON0001166 Email exchange between Anthony de Garr Robinson, Owain Draper and Andrew
Parsons dated 14.07.17
603. I WBONO001167 Draft Suspense Account rider dated 14.07.17
604. I WBON0001168 Email from Amy Prime to Owain Draper and Anthony de Garr Robinson dated
14.07.17
605. I WBONO001171 Email from Andrew Parsons to Jane MacLeod and Mark Underwood dated 14.07.17
606. I POL00024771 Email chain between Mark Underwood, Andrew Parsons and Jane MacLeod dated
14.07.17
607. I WBONO001173 Email exchange involving Andrew Parsons, Mark Underwood, Jane MacLeod, Angela
Van Den Bogerd, Mark Davies, Stuart Nesbit, Patrick Bourke, Rob Houghton, Rodric
AC_207339548_5 30
WITN10390200
WITN10390200
No. I Document Reference Document
Williams, Thomas Moran, Tom Wechsler, Mark Ellis and Melanie Corfield dated
14.07.17
608. I WBON0001176 Email exchange involving Anthony de Garr Robinson, Amy Prime and Owain Draper
dated 14.07.17
609. I WBON0001178 Email exchange involving Mark R Davies, Andrew Parsons, Jane MacLeod, Mark
Underwood, Angela Van Den Bogerd, Mark Ellis, Melanie Corfield, Thomas Moran,
Stuart Nesbitt, Patrick Bourke, Rob Houghton, Tom Wechsler, Kevin Morgan and Amy
Prime dated 11.07.17 — 14.07.17
610. I POL00024489 Email exchange involving Thomas Moran, Andrew Parsons, Mark Underwood, Jane
MacLeod, NGELA Van Den Bogerd, Mark Davis, Stuart Nesbit, Patrick Bourke, Rob
Houghton, Rodric Williams, Tom Wechsler, Mark Ellis, Melanie Corfield dated 11.07.17
— 16.06.17
611. I WBON0000502 Email exchange involving Rob Houghton, Andrew Parsons, Jane MacLeod, Mark
Underwood, Angela Van Den Bogerd, Mark Ellis, Mark R Davies, Melanie Corfield,
Thomas Moran , Stuart Nesbitt, Patrick Bourke, Rob Houghton, Tom Wechsler, Kevin
Morgan and Amy Prime dated 11.07.17 — 17.07.17
612. I POL00024253 Email exchange involving Andrew Parsons, Rob Houghton, Jane MacLeod, Mark
Underwood, Angela Van Den Bogerd, Mark Ellis, Mark R Davies, Melanie Corfield,
Thomas Moran , Stuart Nesbitt, Patrick Bourke, Rob Houghton, Tom Wechsler, Kevin
Morgan and Amy Prime dated 11.07.17 — 17.07.17
613. I WBON0001180 amat fom Andrew Parsons to Anthony de Garr Robinson and Owain Draper dated
614. I WBONO001183 Email from Andrew Parsons to Jane MacLeod and Kevin Morgan dated 17.07.17
615. I WBON0001185 Email Mark Underwood to Andrew Parsons dated 18.07.17
616. I WBON0000468 Email from Mark Westbrook to Andrew Parsons dated 10.10.16
617. I WBON0000469 Draft Project Bramble report dated 07.10.16
618. I WBONO0001064 Email from Jonathan Gribben to Mark Westbrook and Lewis Keating dated 17.10.16
619. I POL00408731 Bond Dickinson Document: Questions on Deloitte's Bramble Draft Report Dated 7
October 2016
620. I WBONO0000472 Email exchange involving Mark Westbrook, Jonathan Gribben, Lewis Keating and
Andrew Parsons dated 10.10.16 — 08.11.16
621. I POL00029104 Bond Dickinson Summary of Deloitte's Bramble report — dated 03.11.16
622. I WBON0001078 Email exchange involving Jonathan Gribben, Mark Westbrook, Lewis Keating, Russell
Norman, Torstein Godeseth and Pete Newsome dated 09.05.17- 11.05.17
623. I WBON0000483 ote chain between Torstein Godeseth, Mark Westbrook and Mark Underwood dated
624. I WBON0001070 ral chain between Jonathan Gribben and Mark Westbrook dated 18.01.17—
AC_207339548_5 4
WITN10390200
WITN10390200
No. I Document Reference Document
625. I WBON0001109 Email from Mark Westbrook to Mark Underwood and Jonathan Gribben dated 08.06.17
626. I POLO0031516 Draft Incomplete Memo outlining conclusions from procedures performed as outlined in
Change Order Number 06 (Version 1) - undated
627. I WBON0001113 Email exchange involving Andrew Parsons, Anthony de Garr Robinson and Owain
Draper dated 19.06.17
628. I POL00174660 Draft Executive Summary — Privileged Users — dated 18.06.17
629. I WBONO001137 Email exchange between Andrew Parsons and Chris Jay dated 04.07.17- 05.07.17
630. I WBONO0001094 Email from Amy Prime to Owain Draper, Anthony de Garr Robinson, Kathryn
Alexander, Huw Williams, Andrew Parsons and Elisa Lukas 07.06.17
631. I WBONO0000482 Email from Anthony de Garr Robinson to Amy Prime and Owain Draper dated
26.06.17
632. I WBONO001119 Email Anthony de Garr Robinson to Andrew Parsons dated 27.06.17
633. I POLO0249555 Draft Generic Defence dated 27.06.17
634. I WBONO001125 Email chain between Amy Prime, Anthony de Garr Robinson and Owain Draper dated
03.07.17
635. I WBON0001124 Email chain between Andrew Parsons and Angela Van Den Bogerd dated 29.06.17
636. I WBONO0000334 Meeting invitation sent by Andrew Parsons to Angela Van Den Bogerd dated 02.07.17
637. I WBONO001126 Email from Andrew Parsons to Anthony de Garr Robinson and Owain Draper dated
03.07.17
638. I WBON0000486 Draft Generic Defence dated 04.07.17
639. I WBONO001141 Email exchange involving Kathryn Alexander, Amy Prime and Shirley Hailstones dated
06.07.17
640. I WBONO001142 Draft Generic Defence — rider for Kathryn Alexander and Shirley Hailstones dated
06.07.17
641. I WBONO001143 Email exchange involving Shirley Hailstones, Kathryn Alexander and Amy Prime
642. I WBON0001144 Draft Generic Defence- dated 06.07.17
643. I POLO0041491 Draft Project Bramble Report
644. I WBON0001192 Email Mark Westbrook to Mark Underwood dated 01.09.17
645. I WBON0001209 Email Mark Westbrook to Jonathan Gribben, Andrew Parsons and Mark Underwood
dated 06.10.17
646. I POL00139454 Bramble Draft Report
647. I WBON0001223 Email from Mark Westbrook to Mark Underwood and Jonathan Gribben dated 02.02.18
AC_207339548_5 32
WITN10390200
WITN10390200
No. I Document Reference Document
648. I POL00139537 Bramble Draft Report
649. I WBON0001318 Email Mark Westbrook to Mark Underwood and Jonathan Gribben dated 12.09.18
650. I POLO0028982 Deloitte Memo — Suspense Document Review dated 18.12.18
651. I POL00139321 Decision Paper: Should Post Office allow Freeths access to Second Sight? for
05.10.16 PLSG meeting
652. I POL00139406 Decision Paper: Should Post Office Allow Freeths access to Second Sight? For
12.07.17 PLSG meeting
653. I POL00250171 Protocol governing Second Sight's interaction with Freeths for the purposes of the
Claim signed by POL 21.08.17
654. I POL00006431 Noting Paper: Update on Litigation Strategy for 16.10.17 PLSG Meeting
655. I POL00357949 Briefing Paper: Electronic Documents Questionnaire for 06.12.17 PLSG Meeting
656. I POL00252428 Decision Paper: Should PO extract all documents from SharePoint for 04.01.17 PLSG
Meeting
657. I POLO0254458 Updating Paper — disclosure for 11.04.18 PLSG meeting
658. I POL00253188 Email Amy Prime to Mark Underwood dated 31.01.18
659 POL00139539 Noting Paper: Update on Strategy for the Court Hearing on 02.02.18
660. I POLO00253355 Email Andrew Parsons to Rodric Williams and Mark Underwood dated 12.02.18
661. I WBON0001226 Email exchange between Andrew Parsons, Rodric Williams and Mark Underwood
dated 12.02.18 to 13.02.18
662. I POL00253363 Email exchange between Amy Prime, Andrew Parsons, Mark Underwood and Rodric
Williams dated 12.02.18 to 13.02.18
663. I WBON0001229 Letter Womble Bond Dickinson to Freeths dated 13.02.18
664. I POL00408810 Summary of Agreed and Disputed Classes of Documents — 19.02.18
665. I POL00253516 POL's Skeleton Argument for 22.02.18 CMC
666. I WBONO0001232 Third CMC Order dated 02.18
667. I POL00022706 Email chain involving Andrew Parsons, Thomas Moran, Jane MacLeod and Mark
Underwood dated 16.04.18
668. I WBON0000177 Email chain involving Andrew Parsons, Thomas Moran, Jane MacLeod and Mark
Underwood dated 16.04.18
669. POL00022705 Email chain involving Andrew Parsons, Thomas Moran, Jane MacLeod and Mark
Underwood dated 16.04.18
670. I POL00254578 Letter from Womble Bond Dickinson to Freeths dated 17.04.18
AC_207339548_5 33
WITN10390200
WITN10390200
No. I Document Reference Document
671. I POLO0285777 Letter Womble Bond Dickinson to Freeths dated 18.05.18
672. I WBON0001690 Letter Womble Bond Dickinson to Freeths dated 18.05.18
673. I POL00285778 Letter Womble Bond Dickinson to Freeths dated 01.08.18
674. I POL00006442 Decision Paper: Should PO disclose the Peak System for 26.09.18 PLSG meeting
675. I WBONO0001566 Seventh CMC Order dated 20.02.19
676. I WBONO001596 Order dated 15.04.19
677. I POLO0023115 CMC Order dated 25.07.19
678. I POLO0285761 Letter Womble Bond Dickinson to Freeths dated 30.08.19
679. I POLO0112568 Update Paper for 26.09.19 PLSG meeting
680. I POLO0286050 Steering Group Paper: Process for Further Issues Claims for meeting on 10.10.19
681. I POLO0241140 Letter Freeths to Rodric Williams dated 28.04.16
682. I POL00408698 Email exchange involving Tom Porter, Andy Garner, Andrew Parsons, Rob Houghton,
Mark Underwood, Shirley Hailstones, and Babu Palathoti dated 04.07.16- 28.07.16
683. I WBON0000460 Email from Andrew Parsons to Pete Newsome dated 19.09.16
684. I WBON0000461 Email chain between Andrew Parsons and Pete Newsome dated 19.09.16 — 21.09.16
685. I WBON0000462 Email chain between Andrew Parsons and Pete Newsome dated 19.09.16- 21.09.16
686. I WBONO0000463 Email exchange involving Andrew Parsons, Pete Newsome and Amy Prime dated
19.09.16 - 23.09.16
687. I WBON0001062 Schedule of Post Office Limited's responses to Claimants’ requests for documents
dated 13.10.16
688. I WBONO0001069 Email exchange involving Andrew Parsons, Pete Newsome, Paul Loraine and Mark
Underwood dated 19.09.16 to 09.01.17
689. I POL00249030 Email exchange involving Rodric Williams, Amanda Pearce and Paul Loraine dated
13.11.16 — 18.11.16
690. I WBON0000473 Email from Paul Loraine to Rodric Williams dated 09.01.17
691. I WBON0000476 Email exchange involving Amy Prime, Andrew Parsons, Paul Loraine and Jonathan
Gribben dated 20.0317 — 21.03.17
692. I POLO0024817 Email from Paul Loraine to Rodric Williams dated 05.04.17
693. I WBONO0001187 Email from Paul Loraine to Rodric Williams and Mark Underwood dated 25.07.17
694. I WBON0001679 Letter Freeths to Bond Dickinson dated 17.03.17
695. I POL00247918 Letter Bond Dickinson to Freeths dated 21.03.17
AC_207339548_5 34
WITN10390200
WITN10390200
No. I Document Reference Document
696. I POL00249567 Email chain between Christopher Jay and Rodric Williams dated 12.06.17 - 30.06.17
697. I WBONO001134 Email chain between Christopher Jay and Rodric Williams dated 12.06.17 -— 30.06.17
698. I WBON0001136 Email exchange involving Andrew Parsons, Pete Newsome, Paul Loraine, Peter
Thompson and Steve Bansal dated 21.09.16- 29.06.17
699. I WBON0000503 Email from Paul Loraine to Andrew Parsons dated 15.08.17
700. I WBONO0000504. Email chain between Pete Newsome and Paul Loraine dated 16.08.17- 17.08.17
701. I WBONO001188 Email from Paul Loraine to Rodric Williams dated 18.08.17
702. I WBON0001196 Letter Bond Dickinson to Freeths dated 01.09.17
703. I WBONO001195 Email chain involving Amy Prime, Anthony de Garr Robinson, Andrew Parsons, Owain
Draper, Peter O'Connell and Megan Atack dated 13.09.17
704. I WBONOO000505 Email exchange involving Anthony de Garr Robinson, Andrew Parsons, Owain Draper,
Amy Prime Elisa Lukas, Megan Atack and Peter O'Connell dated 13.09.17 - 14.09.17
705. I WBONO001197 Email exchange involving Andrew Parsons, Owain Draper, Anthony de Garr
Robinson, Amy Prime and Megan Atack dated 13.09.17- 14.09.17
706. I WBON0001199 Letter Bond Dickinson to Freeths dated 15.09.17
707. POL00041483 Email exchange involving Andrew Parsons, Legal Defence email address at Fujitsu,
Michael Harvey, Pete Newsome and Torstein Godeseth dated 18.09.17 to 21.09.17
708. I WBONO000506 Email exchange Andrew Parsons, Legal Defence email address at Fujitsu, Michael
Harvey, Pete Newsome and Torstein Godeseth dated 18.09.17- 21.09.17
709. I WBONO0001200; Letter Bond Dickinson to Freeths dated 22.09.17
WBONO0001201
710. I WBON0001203 Letter from Freeths to Bond Dickinson dated 27.09.17
711. I WBON0001400 Letter Bond Dickinson to Freeths dated 04.10.17
712. I POL00250828 Email exchange involving Amy Prime, Andrew Parsons, Legal Defence email address
at Fujitsu, Michael Harvey, Pete Newsome and Torstein Godeseth dated 18.09.17—
04.10.17
713. I POLO0250841 Email exchange involving Christopher Jay, Amy Prime, Andrew Parsons, Legal
Defence at Fujitsu, Michael Harvey, Pete Newsome and Torstein Godeseth dated
18.09.17 - 05.10.17
714. I WBON0001213 Letter from Bond Dickinson to Freeths dated 06.10.17
715. I POLO0041458 Email exchange involving Paul Lorraine, Rodric Williams and Miles Trent, dated
26.07.17 — 14.08.17
716. I WBON0000523 Email chain involving Paul Loraine, Mark Underwood, Andrew Parsons and Mandy
Robertson dated 13.05.18
AC_207339548_5 35
WITN10390200
WITN10390200
No. I Document Reference Document
717. I WBONO001687 Email chain involving Mark Underwood, Paul Loraine, Rodric Williams and Jane
MacLeod dated 14.03.18 and 22.03.18
718. I POLO0225914 Correcting Accounts for "Lost" Discrepancies document drafted by Gareth Jenkins
dated 29.09.2010
719. I POL00041040 Email chain involving Pete Newsome, Mark Underwood, Andrew Parsons and Patrick
Bourke dated 08.04.15
720. I WBONO000459 Email from Andrew Parsons to Mark Westbrook dated 15.08.16
721. I WBON0001063 Letter from Bond Dickinson to Freeths dated 13.10.16
722. I WBONO001679 Letter Freeths to Bond Dickinson dated 17.03.17
723. I POL00250437 Protocol Governing Second Sight's Interaction with Freeths for the Purposes of the
Claim 21.08.17 — 01.09.17
724. I WBONO000178 Email from Elisa Lukas to Pete Newsome dated 06.09.17
725. I POL00250090 Letter from Freeths to Bond Dickinson dated 06.07.17
726. I WBONO0000174 Email exchange between Elisa Lukas and Pete Newsome dated 06.09.17 — 08.09.17
727. I WBON0001200 Letter Bond Dickinson to Freeths dated 22.09.18
728. WBONO001207 Email from Anthony de Garr Robinson to Owain Draper; Andrew Parsons, and Amy
Prime dated 05.10.17
729. I WBON0001208 Fourth Witness statement of Andrew Parsons - undated
730. MWBONO0000507 [Email from Elisa Lukas to Pete Newsome dated 03.10.17
731. MWBONO0001684 [Email from Pete Newsome to Elisa Lukas dated 06.10.17
732. IPOL00250836 [Draft Confidentiality Agreement between Fujitsu Services Limited and IT Group Ltd
733. MBONO0000190 [Email chain involving Owain Draper, Anthony de Garr Robinson, Andrew Parsons, Amy
Prime, and Elisa Lukas dated 05.10.17- 07.10.17
734. [POL00252048 Electronic Documents Questionnaire dated 06.12.17
735. )MWBON0001235 Notes of Meeting Between Parties and IT Experts dated 11.04.18
736. IPOL00254961 Letter from Womble Bond Dickinson to Freeths dated 14.05.18
737. [POL00254995 Letter from Freeths to Womble Bond Dickinson dated 18.05.18
738. I[POL00254996 Borparison Between Claimant's and Defendant's Draft Fifth CMC Order dated May
739. )MWBON0001247 [Email exchange between Jonathan Gribben and Pete Newsome dated 21.05.18—
24.05.18
740. MWBONO0000529 Email from Imogen Randall to Jonathan Gribben dated 07.06.18
AC_207339548_5 36
WITN10390200
WITN10390200
No. I Document Reference Document
741. MWBON0000530 Email chain involving Jonathan Gribben, Lucy Bremner, Chris Emery and Andrew
Parsons dated 04.06.18 to 11.06.18
742. MWBONO000531 Email exchange between Jonathan Gribben and Imogen Randall dated 12.06.18
743. MBON0000532 Email exchange between Jonathan Gribben and Imogen Randall dated 12.06.18
744. )WBON0000533 [Email exchange between Jonathan Gribben and Pete Newsome dated 13.06.18
745. MWBON0000534 Email exchange between Jonathan Gribben and Imogen Randall dated 12.06.18 to
14.06.18
746. I[POL00110998 Experts Requests for Information Prepared in Accordance with Paragraph 9 of the Fifth
(CMC Order dated 26.06.18
747. )WBON0001256 [Email exchange between Jonathan Gribben, Robert Worden, Chris Emery and Andrew
Parsons dated 29.06.18 — 30.06.18
748. IPOL00003386 Letter from Freeths to Womble Bond Dickinson dated 02.10.18
749. I WBONO000559 Email chain involving Jason Coyne, Imogen Randall, James Hartley, 1935 Post Office,
Andrew Parsons, Jonathan Gribben, Robert Worden, Anthony de Garr Robinson and
Simon Henderson dated 20.07.18 — 21.07.18
750. MWBONO0001273 [Email exchange between Robert Worden, Andrew Parsons and Jonathan Gribben
dated 31.07.18 - 01.08.18
751. MWBONO001277 Email exchange between Jonathan Gribben, Pete Newsome, Dave Ibbett and Matthew
Lenton dated 01.08.18 — 07.08.18
752. )MWBON0001278 Draft Post Office Response to Mr Coyne's 20.07 email dated 07.08.18
753. MWBONO0001279 Email exchange between Jonathan Gribben, Pete Newsome, Dave Ibbett and Matthew
Lenton dated 01.08.18 — 07.08.18
754. )MWBON0001280 Draft Post Office Response to Mr Coyne's 20.07 email dated 07.08.18
755. NMWBONO0001294 Email exchange between Jonathan Gribben and Pete Newsome dated 13.08.18 —
20.08.18
756. )WBON0001326 Fifth CMC Order dated 25.07.18
757. MWBONO0001668 Letter from Womble Bond Dickinson to Freeths dated 08.08.18
758. I[POL00256155 Post Office's Response to Jason Coyne's "Requests for Information" Document Sent on
(12 July 2018 and dated 26 June 2018
759. IPOL00256731 [Decision paper for 03.09.18 PLSG meeting: Should Post Office apply to strike out the
inadmissible parts of the Claimants’ evidence?
760. IPOL00257086 [Decisions paper for 13.09.18 PLSG meeting: When should Post Office argue its strike
ut application?
761. MWBON0001340 Letter from Womble Bond Dickinson to Freeths dated 25.10.18
AC_207339548_5 37
WITN10390200
WITN10390200
No. I Document Reference Document
762. MWBON0001431 Email exchange between Matthew Lenton, Amy Prime and Lucy Bremner dated
(06.02.19 to 01.03.19
763. WBON0001429 Letter Womble Bond Dickinson to Freeths dated 18.02.19
764. )WBONO0000217 Fujitsu comments on Drop and Go dated 25.03.19
765. )WBON0000211 [Email exchange between Katie Simmonds and Matthew Lenton dated 28.03.19-
29.03.19
766. )WBON0000203 [Email chain involving Katie Simmonds, Matthew Lenton and Amy Prime dated 28.03.19
fF 29.03.19
767. )WBONO000218 [Email from Katie Simmonds to Andrew Parsons dated 29.03.19
768. )WBONO0000276 [Email chain involving Katie Simmonds, Charlie Temperley and Amy Prime dated
29.03.19 — 01.04.19
769. )WBON0000275 Email exchange between Katie Simmonds and Matthew Lenton dated 28.03.19-
(01.04.19
770. MWBONO0001537 Email chain involving Angus McDonald, David Cooke and Katie Simmonds dated
29.03.19 — 02.04.19
771. MWBONO0001535 [Draft Note on Drop and Go Bug dated 02.04.19
772. MWBONO000250 Email chain involving Katie Simmonds, Matthew Lenton and Charlie Temperley dated
28.03.19 — 05.04.19
773. MWBONO000225, [Email exchange between Katie Simmonds, James Brett and David Cooke dated
29.03.19 to 08.05.19
774. MWBONO000274 Email exchange between Katie Simmonds, James Brett and David Cooke dated
29.03.19 to 08.05.19
775. )WBON0001581 [Email exchange between Katie Simmonds, James Brett, David Cooke and Erika
Smithurst dated 29.03.19 to 10.05.19
776. MWBONO0001603 [Email chain involving James Brett, Katie Simmonds, Akshar Vaidya, Henk Bakker and
Maxwell Racher dated 01.08.18 to 10.05.19
777. WWBONO001608 Email from Katie Simmonds to Rebecca Keating and Simon Henderson dated 22.05.19
778. I[POL00132736 Final Version Note on Drop and Go Bug dated 22.05.19
779. )WBON0000227 Email chain involving Katie Simmonds, Rebecca Keating, Simon Henderson and Charlie
[Temperley dated 22.05.19
780. )WBON0000290 [Email from Simon Henderson to Katie Simmonds and Rebecca Keating dated 23.05.19
781. MWBONO000251 Email chain involving Katie Simmonds, Rebecca Keating, Simon Henderson and Charlie
ITemperley dated 22.05.19 — 28.05.19
782. MWBONO000159 [Email from Charlie Temperley to Andrew Parsons dated 28.05.19
AC_207339548_5 38
WITN10390200
WITN10390200
No. I Document Reference Document
783. NMBON0000149 Email from Charlie Temperley to Andrew Parsons dated 29.05.19
784. )WBON0000293 [Email chain involving Amy Prime, Anthony de Garr Robinson, Katie Simmonds, Lucy
[Bremner and Jonathan Gribben dated 17.06.19
785. )MWBON0000138 Email chain involving Amy Prime, Anthony de Garr Robinson, Katie Simmonds, Lucy
[Bremner and Jonathan Gribben dated 17.06.19- 18.06.19
786. )WBON0000139 [Email exchange between Amy Prime, Anthony de Garr Robinson and Owain Draper
dated 17.06.19- 18.06.19
787. IPOL00028062 Deloitte Horizon: Desktop Review of Assurance Sources and Key Control Features:
Draft for Discussion dated 23.05.14
788. )MWBONO0000141 Disclosure Review Briefing Note on Privilege dated 21.01.19
789. MWBONO0000145 (Cartwright King Advice on Disclosure Issues Arising out of the CCRC and Separate Civil
Proceedings drafted by Simon Clarke dated 20.12.17
790. IPOL00027054 Zebra Action Summary drafted by James Rees dated 12.06.14
791. IPOL00002356 Redacted Action Summary drafted by James Rees dated 12.06.14
792. )MWBON0001264 [Email from Amy Prime to Anthony de Garr Robinson and Simon Henderson dated
25.07.18
793. MWBONO000567 Email chain involving Amy Prime, Anthony de Garr Robinson, Simon Henderson and
Andrew Parsons dated 25.07.18 — 26.07.18
794. I[POL00408730 [Email from Amy Prime to Rodric Williams dated 26.07.18
795. IPOL00255961 Email chain involving Amy Prime, Rodric Williams and Mark Underwood dated 26.07.18
fF 27.07.18
796. )WBON0001270 Email chain involving Amy Prime, Anthony de Garr Robinson, Simon Henderson,
Andrew Parsons dated 25.07.18 — 27.07.18
797. MWBON0001339 Letter from Freeths to Womble Bond Dickinson dated 22.10.18
798. )MWBON0000626 [Email from Owain Draper to Amy Prime dated 23.10.18
799. WBON0001362 Letter from Womble Bond Dickinson to Freeths dated 14.11.18
800. )WBON0001364 Letter from Freeths to Womble Bond Dickinson dated 16.11.18
801. MBONO000155 Email chain involving Michael Wharton, Veronica Branton, Kim Pretorius, Amy Prime
land Andrew Parsons dated 19.03.18— 14.05.18
802. )WBONO0000181 [Email chain involving Michael Wharton, Veronica Branton, Kim Pretorius, Amy Prime
land Andrew Parsons dated 19.03.18 — 17.05.18
803. )WBONO0000176 Email chain involving Amy Prime, Dan Cheal and Michael Wharton dated 20.06.18—
28.06.18
804. [POL00266947 Letter Womble Bond Dickinson to Freeths dated 05.03.19
AC_207339548_5 39
WITN10390200
WITN10390200
No. I Document Reference Document
805. MWBONO0001465 Email chain involving Miranda Bond, Charlie Temperley, Andrew Parsons, Amy Prime,
[Emma Campbell-Danesh, Victoria Brooks, Jonathan Gribben, Michael Wharton, Anna
(Martin, Dave Panaech, Lucy Bremner, Beth Hooper, Katie Simmonds, Sushma
MacGeoch, Rachel Lawrie, Jane Atkinson, Mandy Robertson, Simon Henderson and
(Owain Draper dated 05.03.19- 10.03.19
806. )WBON0001449 Letter from Womble Bond Dickinson to Freeths dated 09.03.19
807. )WBONO0001574 [Redacted Document Review dated 18.03.19
808. IPOL00003635 Letter from Womble Bond Dickinson to Freeths dated 27.02.19
809. IPOL00003570 Letter from Womble Bond Dickinson to Freeths dated 15.03.19
810. IPOL00257537 Letter from Freeths to Womble Bond Dickinson dated 19.09.18
811. [POL00285759 Letter from Womble Bond Dickinson to Freeths dated 03.10.18
812. )MWBON0000198 [Email from Lucy Bremner to Mark Hotson dated 24.10.18
813. MBONO0000283 Email chain involving, Lucy Bremner, Mark Hotson, Angela Van-Den-Bogerd, Catherine
Hamilton and Johan Appel dated 24.10.18- 26.10.18
814. )MWBON0001492 Email chain involving Rebecca Reay, Michelle Darbyshire, Rebecca Whipley, Johann
Appel and Lisa Toye dated 31.10.18 - 06.11.18
815. MBONO0000282 Email chain involving Lucy Bremner, Mark Hotson, Angela Van-Den-Bogerd, Catherine
Hamilton, Johann Appel, Mark Underwood and Andrew Parsons dated 24.10.18 —
(02.11.18
816. )WBON0001487 [Email chain involving Lucy Bremner, Mark Hotson, Angela Van-Den-Bogerd, Catherine
Hamilton, Johann Appel, Peter Stanley, Michael Austin and Ben Cooke dated 24.10.18
- 06.11.18
817. I[POL00042127 Email chain involving Rebecca Reay, Michelle Darbyshire, Johann Appel and Lucy
[Bremner dated 31.10.18 — 09.11.18
818. MBONO0001359 Email chain involving Rebecca Reay to Michelle Darbyshire, Johann Appel, Lucy
[Bremner and Rodric Williams dated 31.10.18 — 09.11.18
819. )WBON0001360 [Email chain involving Rodric Williams, Luke Ryan and Lucy Bremner dated 12.11.18
820. MWBONO001693 Email from Rodric Williams to Lucy Bremner dated 12.11.18
821. )MWBONO0001485 Email chain involving Rodric Williams, Luke Ryan, Johann Appel and Andrew Parsons
dated 12.11.18 — 15.03.19
822. MWBON0001412 Letter from Freeths to Womble Bond Dickinson dated 08.02.19
823. [POL00263874 Letter from Womble Bond Dickinson to Freeths dated 11.02.19
824. )WBON0001416 Letter from Womble Bond Dickinson to Freeths dated 12.02.19
825. MWBONO0000170 [Email exchange between Lucy Bremner and Johann Appel dated 11.02.19- 12.02.19
AC_207339548_5 40
WITN10390200
WITN10390200
No. I Document Reference Document
826. )WBON0000173 Email exchange between Lucy Bremner and Johann Appel dated 11.02.19- 13.02.19
827. )MWBON0000167 [Email exchange between Lucy Bremner and Johann Appel dated 11.02.19 — 13.02.19
828. )WBONO0001472 Letter from Freeths to Womble Bond Dickinson dated 13.03.19
829. )MWBON0000320 [Email chain involving Angelique Richardson, Amy Prime, Andrew Parsons, Emma
(Campbell-Danesh, Victoria Brooks, Jonathan Gribben, Michael Wharton, Anna Martin,
[Dave Panaech, Lucy Bremner, Beth Hooper, Katie Simmonds, Sushma MacGeoch,
Rachel Lawrie, Jane Atkinson and Charlie Temperley dated 13.03.19
830. WBON0000322 Draft Letter from Womble Bond Dickinson to Freeths dated 13.03.19
831. MBON0001483 [Email from Amy Prime to Andrew Parsons and Rodric Williams dated 14.03.19
832. MBONO000201 [Email from Rodric Williams to Luke Ryan dated 14.03.19
833. IPOL00022691 Letter from Womble Bond Dickinson to Luke Ryan dated 14.03.19
834. MBON0001488 Prva chan involving Rodric Williams, Luke Ryan and Andrew Parsons dated 12.11.18
835. IPOL00042127 Email chain involving Rebecca Reay (Whibley), Michelle Darbyshire, Johann Appel and
Lucy Bremner dated 31.10.18 — 09.11.18
836. )WBONO0001491 Email chain involving Rebecca Reay, Michelle Darbyshire, Johann Appel and Lisa Toye
dated 31.10.18 - 13.11.18
837. MWBON0001487 [Email chain involving Lucy Bremner, Mark Hotson, Angela Van-Den-Bogerd, Catherine
Hamilton, Johann Appel, Peter Stanley, Michael Austin and Ben Cooke dated 24.10.18
- 06.11.18
838. [POL00269022 Letter from Womble Bond Dickinson to Freeths dated 15.03.19
839. )WBONO0001510 Letter from Royal Mail Group to Freeths
840. [POL00269053 [Thirteenth Witness Statement of Andrew Paul Parsons dated 21.03.19
841. )WBONO0001682 (Claimants' Response to the Defendant's Request For Further Information Under CPR
Part 18 Dated 27 April 2017
842. )WBON0001220 Letter Womble Bond Dickinson to Freeths dated 01.11.17
843. WBONO0001219 Letter Freeths to Womble Bond Dickinson dated 09.11.17
844. )WBON0001337 [Transcript of CMC on 22.02.18
845. [POL00110872 (Claimants' Proposed Factual Matrix dated 02.03.18
846. [POL00363651 Letter Womble Bond Dickinson to Freeths dated 20.03.18
847. [POL00006408 Briefing Paper: Witness statements — how to deal with inadmissible evidence — for
IPLSG meeting 28.03.18
848. [POL00041899 \Transcript of CMC on 05.06.18
AC_207339548_5 4
WITN10390200
WITN10390200
No. I Document Reference Document
849. [POL00358103 [Decision paper: Should Post Office respond to inadmissible allegations in its witness
evidence? — for PLSG meeting 17.07.18
850. [POL00255849 Letter Womble Bond Dickinson to Freeths dated 19.07.18
851. [POL00255848 Letter Freeths to Womble Bond Dickinson dated 19.07.18
852. [POL00006455 [Decision Paper: Should Post Office defer its strike out application to the start of trial?—
for PLSG meeting 13.09.18
853. WBON0001313 Letter Womble Bond Dickinson to Freeths dated 31.08.18
854. IPOL00256731 [Decision Paper: Should Post Office apply to strike out inadmissible parts of the
(Claimants' evidence? — for PLSG meeting 03.09.18
855. I[POL00256627 Note on admissibility of evidence for the Common Issues Trial— undated
856. [POL00256583 [Draft Ninth Witness Statement of Andrew Paul Parsons — undated
857. )WBON0001700 Decision paper on strike out
858. MBON0001254 Email exchange involving Gideon Cohen, Andrew Parsons, Anthony de Garr Robinson,
[David Cavender, Simon Henderson, Owain Draper, Imogen Randall, Chloe Squibb and
Amy Prime dated 22.06.18
859. )MWBONO0000535, Email exchange involving Victoria Brooks, Andrew Parsons, Gideon Cohen, Anthony de
(Garr Robinson, David Cavender, Simon Henderson and Owain Draper dated 22.06.18
fo 23.06.18
860. )WBONO0000536 [Email chain involving Victoria Brooks, Mandy Robertson and Lucy Garland dated
26.06.18
861. )WBON0000540 [Email from Victoria Brooks to Ed Duffield, Helen Creech, Ivan Roots, Dave Panaech
jand Mandy Robertson dated 03.07.18
862. MWBONO0000568 [Email Victoria Brooks to Andrew Parsons dated 27.07.18
863. )WBONO0000609 [Email Victoria Brooks to Andrew Parsons dated 15.08.18
864. WBONO0000569 [Email Victoria Brooks to Andrew Parsons dated 31.07.18
865. MBONO0000582 Email exchange between Victoria Brooks. Andrew Parsons and Amy Prime dated
(01.08.18 to 02.08.18
866. )WBONO000608 Email chain involving Victoria Brooks. Helen Creech, Ed Duffield, Andrew Parsons,
(Gideon Cohen and Owain Draper dated 14.08.18
867. )WBON0001271 Email Mandy Robertson to Paul Williams dated 27.07.18
868. )WBON0000583 [Email from Victoria Brooks to Andrew Parsons dated 03.08.18
869. MBONO000606 Email exchange between Andrew Parsons, Mandy Robertson and Victoria Brooks dated
(09.08.18 to 13.08.18
870. )WBONO0001281 [Email Andrew Parsons to Paul Williams dated 14.08.18
AC_207339548_5 42
WITN10390200
WITN10390200
No. I Document Reference Document
871. )MWBON0001289 Email chain involving Andrew Parsons, Paul Williams and Mandy Robertson dated
14.08.18 to 16.04.18
872. MWBONO0001287 Email chain involving Andrew Parsons, Angela van den Bogerd, Rodric Williams and
Paul Williams dated 14.08.18 to 16.08.16
873. IPOL00041921 Email chain involving Andrew Parsons, Angela van den Bogerd, Rodric Williams and
Paul Williams dated 14.08.18 to 16.08.16
874. )MWBON0001257 Email from Victoria Brooks to David Longbottom dated 25.07.18
875. WBON0000610 [Email from Andrew Parsons to Victoria Brooks dated 20.08.18
876. )WBONO0000612 [Email Victoria Brooks to Andrew Parsons dated 21.08.18
877. )WBON0001295 [Email exchange between Andrew Parsons and David Longbottom dated 21.08.18
878. MBONO0000618 Email exchange between Victoria Brooks and Andrew Parsons dated 22.08.18
879. )WBON0000619 [Email chain involving Andrew Parsons, Mandy Robertson and Rodric Williams dated
22.08.18
880. )WBONO0000517 Notes of Meeting Between Victoria Brooks. Mandy Robertson and Angela van den
Bogerd dated 15.01.18
881. MWBONO0000515 Notes of Meeting Between Victoria Brooks. Mandy Robertson and Angela van den
Bogerd dated 15.01.18
882. )WBONO0000546 Proof of evidence of Angela van den Bogerd — 22.05.18
883. MBONO000556 Email lvan Roots to Angela van den Bogerd dated 06.07.18
884. NMBONO0000560 [Email exchange between Ivan Roots and Angela van den Bogerd dated 06.07.18 to
23.07.18
885. )WBON0000557 [Email Ivan Roots to Victoria Brooks dated 12.07.18
886. MWBONO0000562 Email chain involving Victoria Brooks, Andrew Parsons, Ivan Roots and Angela van den
Bogerd dated 06.07.18 to 24.07.18
887. )WBONO0000580 [Email Ivan Roots to Andrew Parsons dated 01.08.18
888. MBON0001286 [Email Andrew Parsons to Angela van den Bogerd dated 08.08.18
889. )MWBONO0000332 Email chain involving Andrew Parsons, Owain Draper, Gideon Cohen, Angela van den
Bogerd and Rodric Williams dated 08.08.18
890. I[POL00363477 Email chain involving Andrew Parsons, David Cavender, Angela van den Bogerd and
Rodric Williams dated 08.08.18 to 20.08.18
891. MBONO000616 Email exchange between Ivan Roots and Andrew Parsons dated 22.08.18
892. I[POL00041956 Draft Witness Statement of Angela Van Den Bogerd dated 23.07.18
AC_207339548_5 43
WITN10390200
WITN10390200
No. I Document Reference Document
893. IPOL00041955 Email chain involving Angela Van Den Bogerd, Andrew Parsons and Rodric Williams
dated 08.08.18 — 20.08.18
894. IPOL00363491 Email Andrew Parsons to Owain Draper and Gideon Cohen dated 22.08.18
895. [POL00363453 Draft witness statement of Angela van den Bogerd dated 10.08.18
896. IPOL00363501 Draft witness statement of Angela van den Bogerd dated 23.08.18
897. MBONO0000625 Email chain involving Angela van den Bogerd, Rodric Williams, Andrew Parsons, David
(Cavender and Gideon Cohen dated 08.08.18 to 23.08.18
898. WBONO0001311 Email chain involving Angela van den Bogerd, Rodric Williams, Andrew Parsons, David
(Cavender, Gideon Cohen and Owain Draper dated 08.08.18 to 23.08.18
899. [POL00363552 [Email Andrew Parsons to Angela van den Bogerd dated 24.08.18
900. )WBON0001274 Email exchange between Paul Williams and Mandy Robertson dated 30.07.18 to
(03.08.18
901. )WBONO001291 Email exchange between Mandy Robertson and Mike Webb dated 17.08.18
902. WBONO0001296 [Email James Cox to Tim Dance dated 22.08.18
903. )WBON0001299 [Email Helen Creech to Helen Dickinson dated 22.08.18
904. )WBONO0000623 [Email lvan Roots to Michael Shields dated 23.08.18
905. )WBON0001301 [Email exchange between Mandy Robertson and Michael Haworth dated 27.07.18 to
23.08.18
906. )WBON0001307 Email Beth Hooper to David Longbottom dated 23.08.18
907. IPOL00154271 Information for Witnesses document — 10.07.18
908. MWBONO0000635, Email chain between Gideon Cogen, Dave Panaech, Mandy Robertson, Amy Prime,
(Andrew Parsons and Victoria Brooks dated 13.11.18
909. MBONO0000636 Email chain between Gideon Cogen, Dave Panaech, Mandy Robertson, Amy Prime,
Andrew Parsons and Victoria Brooks dated 13.11.18
910. MBONO0000637 Email chain between Owain Draper, Gideon Cogen, Dave Panaech, Mandy Robertson,
Amy Prime, Andrew Parsons and Victoria Brooks dated 13.11.18
911. MWBONO0001379 Email chain between Amy Prime, Owain Draper and David Cavender dated 12.12.18
912. MBON0000642 Email Amy Prime to Imogen Randall dated 14.12.18
913. WBON0000643 Flowchart 1 Transaction Corrections — 13.12.18
914. WBONO0001393 Letter Freeths to Womble Bond Dickinson dated 17.12.18
915. MWBONO0000645 [Email chain involving Amy Prime, Owain Draper, Gideon Cohen, David Cavender,
[Stephanie Jameson, Andrew Parsons, Jonathan Gribben, Emma CampbeltDanesh,
AC_207339548_5 44
WITN10390200
WITN10390200
No. I Document Reference Document
ictoria Brooks, Michael Wharton, Anna Martin, Dave Panaech, Lucy Bremner, Beth
'Hooper, Katie Simmonds and Charlie Temperley dated 17.12.18
916. MWBONO0000646 Email chain involving Amy Prime, Owain Draper, Gideon Cohen, David Cavender,
Stephanie Jameson, Andrew Parsons, Jonathan Gribben, Emma CampbeltDanesh,
‘ictoria Brooks, Michael Wharton, Anna Martin, Dave Panaech, Lucy Bremner, Beth
Hooper, Katie Simmonds and Charlie Temperley dated 17.12.18
917. MWBONO0001396 [Email chain involving Amy Prime, Owain Draper, Gideon Cohen, David Cavender,
Stephanie Jameson, Andrew Parsons, Jonathan Gribben, Emma CampbeltDanesh,
ictoria Brooks, Michael Wharton, Anna Martin, Dave Panaech, Lucy Bremner, Beth
Hooper, Katie Simmonds and Charlie Temperley dated 17.12.18
918. MWBON0001328 Emails between Amy Prime and David Cavender dated 17.09.18 to 21.09.18
919. )WBON0001331 [Emails between Amy Prime and David Cavender dated 17.09.18 to 21.09.18
920. IPOL00257368 Updated First round of Evidence for the Horizon Trial dated 25.09.18
921. [POL00257886 \Update Paper: Supplemental Evidence for the Horizon Trial for meeting on 12.10.18
922. )WBONO0001269 Fifth CMC Order dated 24.07.18
923. [POL00358213 (Claimants Outline Document in relation to the Horizon issues dated 17.08.18
924. IPOL00258234 Expert Report of Jason Coyne dated 16.10.18
925. IWBONO000632 Horizon Issues Trial Witness Evidence Plan
926. MBONO0000629 Email exchange between Anthony de Garr Robinson, Jonathan Gribben and Simon
[Henderson dated 18.10.18 — 30.10.18
927. )WBON0000630 [Email exchange between Anthony de Garr Robinson, Jonathan Gribben and Simon
[Henderson dated 18.10.18 — 31.10.18
928. )MBONO0000627 Email exchange between Jonathan Gribben and Andrew Parsons dated 30.10.18
929. )WBON0001694 [Email exchange between Jonathan Gribben, Simon Henderson and Anthony de Garr
[Robertson dated 12.11.18 to 15.11.18
930. IPOLO0111481 Expert Report of Dr Robert Worden dated 07.12.18
931. [POL00262929 Supplemental Expert Report of Jason Coyne dated 01.02.19
932. IPOLO0266866 Second Joint Statement of Experts dated 25.02.19
933. I[POL00278807 Post Office's Written closing Submissions: Horizon Issue Trial— dated 27.06.19
934. IPOL00026918 [Third Joint Statement of Experts dated 01.03.19
935. )MWBON0001723 Simon Clarke of Cartwright King's advice on the use of expert evidence relating to the
integrity of the Fujitsu Services Ltd Horizon System dated 15.07.13
936. MBONO001005 Email from Andrew Parsons to Anthony de Garr Robinson dated 01.06.16
937. )MWBONO0001011 {Commentary on Documents in Counsel's Bundle dated 31.05.16
AC_207339548_5 45
WITN10390200
WITN10390200
No. I Document Reference Document
938. MBONO0001315 Email chain involving Andrew Parsons, Anthony de Garr Robinson, Simon Henderson
land Robert Worden dated 06.09.18 - 07.09.18
939. [POL00042010 [Email Andrew Parsons to Rodric Williams dated 07.09.18
940. IWBON0000342 Email exchange between Anthony de Garr Robinson, Jonathan Gribben and Simon
[Henderson dated 12.11.18
941. )WBON0000341 [Email exchange between Anthony de Garr Robinson, Jonathan Gribben and Simon
[Henderson dated 12.11.18
942. MBONO0000189 Email exchange between Anthony de Garr Robinson, Andrew Parsons and Simon
[Henderson dated 16.11.18— 18.11.18
943. MBONO0000331 [Email from Anthony de Garr Robinson to Jonathan Gribben, Andrew Parsons,
Sebastian Isaac, Lucy Bremmer and Katie Simmons dated 07.12.18
944. MWBONO001721 Email from Anthony de Garr Robinson to Jonathan Gribben, Andrew Parsons,
Sebastian Isaac, Lucy Bremmer and Katie Simmons dated 07.12.18
945. IPOL00111371 [Email chain involving Andrew Parsons, Gareth Jenkins, Jonathan Gribben, Matthew
Lenton, Pete Newsome and Katie Simmonds dated 16.11.18
946. )WBONO0000196 [Email Andrew Parsons to Katie Simmonds dated 16.11.18
947. )WBON0000197 Second Witness Statement of Angela Margaret Van Den Bogerd — draft dated 16.11.18
948. )MWBON0000287 Email exchange between Katie Simmonds and Andrew Parsons dated 16.11.18
949. )WBON0000285 [Email exchange between Katie Simmonds and Andrew Parsons dated 16.11.18
950. IBONO000286 Second Witness Statement of Angela Margaret Van Den Bogerd- draft dated 16.11.18
951. )MBON0000195 Email exchange between Angela van den Bogerd and Katie Simmonds dated 16.11.18
952. MWBON0000292 [Email from Simon Henderson to Andrew Parsons and Jonathan Gribben dated 12.02.19
953. MWBON0001432 [Email chain involving Angela Van-Den-Bogerd, Emma Campbell-Danesh, Anthony de
(Garr Robinson, Simon Henderson, Owain Draper, Andrew Parsons and Paul I Smith
dated 11.02.19- 04.03.19
954. MBONO0000288 Email chain involving Steve Parker, Dave Ibbett, Pete Newsome, Jonathan Gribben and
Andrew Parsons dated 22.10.18 to 06.11.18
955. )WBON0000192 [Email from Jonathan Gribben to Pete Newsome dated 15.10.18
956. )MWBON0000193 Response to Richard Roll (To be turned into a Statement by Steve Parker) dated
(15.10.18- undated
957. MWBONO0000194 Email chain involving Steve Parker, Jonathan Gribben and Pete Newsome dated
(15.10.18 — 16.10.18
AC_207339548_5 46
WITN10390200
WITN10390200
No. I Document Reference Document
958. IPOL00258674 Briefing for Paula Vennells — Response to Richard Roll's evidence dated 30.10.18
959. )WBON0000284 [Email exchange between Andrew Parsons and Rodric Williams dated 30.10.18—
31.10.18
960. [POL00266514 Second Witness Statement of Stephen Paul Parker dated 29.01.19
961. )WBON0001218 [Third Witness Statement of Torstein Olav Godeseth dated 28.02.19
962. )WBON0001401 Email exchange between Jonathan Gribben and Steve Parker dated 29.01.19 —
29.01.19
963. )WBON0001402 [Email chain involving Lucy Bremner, Steve Parker and Jonathan Gribben dated
29.01.19
964. [POL00363893 [Email chain involving Lucy Bremner, Steve Parker and Jonathan Gribben dated
29.01.19
965. MBONO0000168 Email chain involving Gareth Jenkins, Pete Newson, Christopher Jay,
Legal. Defence@uk.fujitsu.com, Jonathan Gribben and Steve Parker dated 29.01.19-
(30.01.19
966. MBONO000202 Email chain involving Jonathan Gribben. Matthew Lenton, Dave Ibbett, Pete Newsome
jand Gareth Jenkins dated 21.01.19 to 07.03.19
967. )WBONO000210 [Email chain involving Matthew Lenton, Jonathan Gribben, Steve Parker, Dave Ibbett,
[Pete Newsome and Gareth Jenkins dated 21.01.19- 11.03.19
968. )WBON0001473 (Corrections to the Second Witness Statement of Stephen Paul Parker dated 14.03.19
969. IPOL00006471 Steering Group Meeting — Noting Paper Expert Report of Dr Robert Worden dated
28.11.18
970. )WBONO000701 [Email from Robert Worden to Andrew Parsons and Jonathan Gribben dated 27.03.19
971. IWBONO000702 [Email chain involving Anthony de Garr Robinson, Andrew Parsons, Simon Henderson,
(Owain Draper, Rebecca Keating, and Robert Worden dated 27.03.19-— 28.03.19
972. )MWBON0001536 Email chain involving Owain Draper, Anthony de Garr Robinson, Andrew Parsons and
[Robert Worden dated 27.03.19 - 28.03.19
973. IPOL00042611 [Email from Andrew Parsons to Rodric Williams dated 06.04.19
974. IPOLO0112051 Email chain between Andrew Parsons, Robert Worden and Jonathan Gribben dated
28.03.19
975. [POL00042614 [Email from Robert Worden to Andrew Parsons dated 06.04.19
976. )WBON0000703 Email from Andrew Parsons to Anthony de Garr Robinson dated 07.04.19
977. )WBONO000704 [Email exchange between Anthony de Garr Robinson and Andrew Parsons dated
(07.04.19 — 08.04.19
978. )WBON0001539 [Email chain involving Owain Draper, Anthony de Garr Robinson and Andrew Parsons
(dated 07.04.19 - 08.04.19
AC_207339548_5 47
WITN10390200
WITN10390200
No. I Document Reference Document
979. )WBON0001543 Letter from Womble Bond Dickinson to Freeths dated 10.04.19
980. )WBONO0001547 [Email chain involving Rodric Williams, Andrew Parsons and Jonathan Gribben dated
(10.04.19
981. IPOL00112145 Email chain involving Andrew Parsons, Robert Worden, Jonathan Gribben, Jason
(Coyne and Siobhan Forster dated 10.04.19 - 16.04.19
982. )WBONO000708 [Email chain involving Robert Worden, Andrew Parsons, Jonathan Gribben, Jason
(Coyne and Siobhan Forster dated 25.04.19
983. )WBON0000710 [Email from Robert Worden to Andrew Parsons dated 26.04.19
984. MBONO000711 [Email chain involving Andrew Parsons, Anthony de Garr Robinson, Simon Henderson
jand Robert Worden dated 26.04.19
985. IWBONO0000712 Email chain involving Anthony de Garr Robinson, Andrew Parsons, Simon Henderson
land Robert Worden dated 26.04.19
986. )WBONO0001578 [Email from Andrew Parsons to Rodric Williams dated 30.04.19
987. [POL00274897 Letter from Womble Bond Dickinson to Freeths dated 03.05.19
988. IPOL00274899 Letter from Freeths to Womble Bond Dickinson dated 07.05.19
989. MWBONO001585 Email chain involving Andrew Parsons, Anthony de Garr Robinson, Simon Henderson
Alan Watts, Rodric Williams. Kirsten Massey and Tom Henderson dated 16.05.19
990. MWBONO0001590 [Email chain involving Andrew Parsons, Anthony de Garr Robinson, Simon Henderson,
Alan Watts, Rodric Williams. Kirsten Massey and Tom Henderson dated 16.05.19
991. MBONO001600 Email chain involving Andrew Parsons, Anthony de Garr Robinson, Simon Henderson,
Kirsten Massey, Alan Watts, Rodric Williams and Tom Henderson dated 16.05.19-—
22.05.19
992. [POL00042688 [Email chain involving Rodric Williams, Kirsten Massey, Andrew Parsons, Alan Watts
jand Tom Henderson dated 16.05.19 — 22.05.19
993. MBONO0000714 [Email chain involving Robert Worden, Jason Coyne, James Hartley and Angela Fraser
dated 22.05.19
994. )WBONO001607 Email chain involving Andrew Parsons, Robert Worden, Katie Simmonds, Jonathan
(Gribben and Lucy Bremner dated 21.05.19 — 22.05.19
995. IPOL00112279 [Email chain involving Robert Worden, Andrew Parsons, Patrick Green, Angela Fraser,
Angelique Richardson, Anthony de Garr Robinson and David Cavender dated 22.05.19
996. [POL00275716 Seventeenth Witness Statement of Andrew Paul Parsons dated 31.05.19
997. )WBONO0001633 Letter from Womble Bond Dickinson to Freeths dated 28.05.19
998. )WBONO0001640 Letter from Womble Bond Dickinson to Freeths dated 03.06.19
999. )WBONO000713 [Email chain involving Andrew Parsons, Anthony de Garr Robinson, Jonathan Gribben,
Simon Henderson and Rodric Williams dated 30.04.19 — 01.05.19
AC_207339548_5 48
WITN10390200
WITN10390200
No. I Document Reference Document
1000. IPOL00042714 \Transcript of Horizon Issues Trial Day 13 dated 23.05.19
1001. IPOL00006753 Minutes of Sub-Committee Meeting 21.02.19
1002. I WBON0000337 Email from Andrew Parsons to Anthony de Garr Robinson dated 15.02.19
1003. WBONO001421 [Email exchange between Anthony de Garr Robinson and Andrew Parsons dated
(15.02.19 - 17.02.19
1004. POL00112903 Speaking Note for Board Sub-Committee dated 21.02.19
1005. WBON0001418 [Email exchange between Anthony de Garr Robinson and Andrew Parsons dated
(15.02.19
1006. JPOL00024150 [Email from Jane MacLeod to Tim Parker, Ken McCall, Tom Cooper and Alisdair
(Cameron dated 20.02.19
1007. POLO0265865 Horizon Issues Trial — Draft Risk Assessment table dated 18.02.19
1008. (WBON0001422 Speaking Note for Board Sub-committee dated 21.02.19
1009. [POL00022940 Initial Summary of the Common Issues Judgment dated 09.03.18
1010. IWBONO001446 [Email exchange involving Mark Underwood, Jane MacLeod, Mark Davies, Melanie
(Corfield, Julie Thomas, Zoe Brauer, Ben Beabey, Andrew Parsons, David Cavender,
(Gideon Cohen, Dave Panaech and Amy Prime dated 08.03.19
1011. POL00267481 [Table of actions following receipt of Common Issues Judgment — 08.03.19
1012. WBONO001366 Post Office's Written Closing Submissions Common Issues Trial- 03.12.18
1013. WBONO001463 [Email from Andrew Parsons to David Cavender and Gideon Cohen dated 09.03.19
1014. WBONO000650 [Email exchange between Gideon Cohen, Andrew Parsons and David Cavender dated
10.03.19
1015. I WBONO0000649 Email exchange between David Cavender, Andrew Parsons and Gideon Cohen dated
(09.03.19
1016. IWBONO001466 [Email chain involving David Cavender, Andrew Parsons and Gideon Cohen dated
(09.03.19 — 10.03.19
1017. POL00267565 [David Cavender QC Advice Note dated 10.03.19
1018. WWBONO000205 Email exchange involving Mark Underwood, Mark Davies, Melanie Corfield, Julie
Thomas, Zoe Brauer, Ben Beabey, Rodric Williams, Jane MacLeod and Andrew
Parsons dated 08.03.19 to 10.03.19
1019. WBONO000652 Email from Amy Prime to Tom Beezer and Andrew Parsons dated 11.03.19
1020. WWBONO000653 Email exchange between Amy Prime, Tom Beezer and Andrew Parsons dated 11.03.19
1021. IWWBONO000654 [Email from Amy Prime to Mark Underwood, Jane MacLeod and Rodric Williams dated
(11.03.19
AC_207339548_5 49
WITN10390200
WITN10390200
No. I Document Reference Document
1022. WWBONO000655 Email chain involving Tom Beezer, Andrew Parsons, Jane MacLeod, Mark Underwood,
Amy Prime and Rodric Williams dated 11.03.19 - 12.03.19
1023. WWBONOO00656 Email chain involving Anthony de Garr Robinson, Andrew Parsons and David Cavender
dated 11.03.19 — 12.03.19
1024. WBONO000657 [Email chain between Adam Sloane and Andrew Parsons dated 12.03.19
1025. WBONO000658 [Email from Tom Beezer to Jane MacLeod dated 12.03.19
1026. IPOL00023930 Email from Amy Prime to Rob Smith dated 12.03.19
1027. POL00023988 [Email from Tom Beezer to Jane MacLeod dated 12.03.19
1028. WBONO000660 Email exchange between Luke Carvalho and Tom Beezer dated 13.03.19
1029. WBONOO00665 [Email exchange between Luke Carvalho and Tom Beezer dated 13.03.19- 15.03.19
1030. WBONO001468 [Email exchange between Amy Prime and Rob Smith dated 12.03.19
1031. WBONO001469 Index to Bundle to Instructions to Counsel for Appeal dated 12.03.19
1032. WBON0001470 [Email exchange between Amy Prime and Rob Smith dated 12.03.19
1033. WBONO000659 Email exchange between Amy Prime and Rob Smith dated 12.03.19
1034. WBON0001474 [Email exchange between Amy Prime and Rob Smith dated 12.03.19- 14.03.19
1035. IPOL00371317 Note on background to possible recusal application dated 13.03.19
1036. POL00025910 (Observations on Recusal Application 14.03.19
1037. IPOL00023898 [Email chain involving Jane MacLeod, Tom Beezer, Andrew Parsons, Tim Parker and
[Thomas Cooper dated 15.03.19
1038. IPOL00023955 Recusal Note - undated
1039. (WBON0001493 [Email chain involving Tom Beezer, David Cavender and Andrew Parsons dated
(15.03.19
1040. WBONO001494 [Draft Recusal Note — 15.03.19
1041. WBONO001495 [Email chain involving Tom Beezer, David Cavender and Andrew Parsons dated
15.03.19
1042. WBONO001496 [Draft Recusal Note — 15.03.19
1043. (WBON0001497 Email chain involving Tom Beezer, David Cavender and Andrew parsons dated
(15.03.19 - 16.03.19
1044. POL00022960 Email chain involving Tom Beezer, Jane MacLeod, Andrew Parsons, Tim Parker and
IThomas Cooper dated 15.03.19- 16.03.19
1045. WBONO001499 Email chain involving Tom Beezer, David Cavender and Andrew Parsons dated
(15.03.19 — 16.03.19
AC_207339548_5 50
WITN10390200
WITN10390200
No. I Document Reference Document
1046. WBONO001500 [Draft Recusal Note dated 16.03.16
1047. I WBONO000666 [Email chain involving Tom Beezer, David Cavender and Andrew Parsons dated
(15.03.19 — 16.03.19
1048. I WBONO000667 Email chain involving Amy Prime, Andrew Parsons, Tom Beezer and David Cavender
ated 15.03.19 - 16.03.19
1049. WBONO000668 [Draft Recusal Note dated 16.03.16
1050. IPOL00023911 [Email chain involving Tom Beezer, Jane MacLeod, Andrew Parsons, Tim Parker and
[Thomas Cooper dated 15.03.19 - 16.03.19
1051. POL00268458 [Draft Recusal Note dated 16.03.16
1052. IPOL00268459 [Draft Recusal Note Comparison Document dated 16.03.16
1053. JPOL00330036 Email chain involving Tom Beezer, Jane MacLeod, Andrew Parsons, Tim Parker and
Thomas Cooper dated 15.03.19- 16.03.19
1054. POL00023231 Email chain involving Tom Beezer, Jane MacLeod, Andrew Parsons, Tim Parker and
[Thomas Cooper dated 15.03.19 — 16.03.19
1055. IPOL00268479 [Draft Recusal Note dated 16.03.16
1056. IPOL00023229 Email chain involving Jane MacLeod, Tom Beezer, Andrew Parsons, Tim Parker and
[Thomas Cooper dated 15.03.19 - 17.03.19
1057. IPOL00268503 [Document prepared by Jane MacLeod: "Extracts from the Judgment that Demonstrate
Bias" — 17.03.19
1058. POL00022969 Email chain involving Tom Beezer, Jane MacLeod, Andrew Parsons, Tim Parker and
[Thomas Cooper dated 15.03.19— 17.03.19
1059. POL00268516 [Draft Recusal Note — 17.03.19
1060. WBONO000672 Email chain involving Jane MacLeod, Tom Beezer, Andrew Parsons, Tim Parker and
[Thomas Cooper dated 15.03.19— 17.03.19
1061. I WBON0000673 [Email chain involving Tom Beezer, Jane MacLeod, Andrew Parsons, Tim Parker and
[Thomas Cooper dated 15.03.19— 17.03.19
1062. WBONO001501 Email chain involving Tom Beezer, Rob Smith and Anthony Grabiner dated 17.03.19
1063. POL00268533 [Draft Recusal Note — 17.03.19
1064. WBONO0000671 [Email from Tom Beezer to Rob Smith dated 17.03.19
1065. WBONO000675 [Email chain involving Jane MacLeod, Tom Beezer, Andrew Parsons, Tim Parker and
IThomas Cooper dated 15.03.19 — 17.03.19
1066. WBON0001351 itness statement of John Breeden dated 14.08.18
1067. POL00268834 Note of Conference with Lord Grabiner QC dated 18.03.19
AC_207339548_5 51
WITN10390200
WITN10390200
No. I Document Reference Document
1068. [POL00022883 Email chain involving Tom Beezer, Jane MacLeod, Richard Watson, Rodric Williams
jand Thomas Cooper dated 18.03.19 — 20.03.19
1069. POL00330038 Email chain involving Tom Beezer, Jane MacLeod, Richard Watson, Rodric Williams
jand Thomas Cooper dated 18.03.19 — 20.03.19
1070. IPOL00021562 Minutes of Board Meeting dated 18.03.19
1071. WBONO000677 [Email from Jane MacLeod to Tom Beezer, Rodric Williams and Andrew Parsons dated
(18.03.19
1072. WBONO001511 Email exchange between Jane MacLeod, Tom Beezer, Rodric Williams and Andrew
Parsons dated 18.03.19 to 19.03.19
1073. WBONO001714 [Email Jane MacLeod to Tom Beezer dated 20.03.19
1074. IWBONO000679 Email chain involving Andrew Parsons, Tom Beezer, Gideon Cohen, David Cavender,
jStephanie Wood and Jane MacLeod dated 20.03.19
1075. WBONO0000681 [Email chain involving David Cavender, Tom Beezer, Andrew Parsons, Gideon Cohen,
Stephanie Wood and Jane MacLeod dated 20.03.19
1076. IWBONO000682 Email exchange between Tom Beezer and Jane MacLeod dated 20.03.1942
1077. WBONO001515 [Email chain involving Tom Beezer, Andrew Parsons, Gideon Cohen, David Cavender,
Stephanie Wood and Jane MacLeod dated 20.03.19
1078. IPOL00269796 Updated Note of Post Office Board Diakin attended by Lord Grabiner (by phone) of
20.03.19
1079. WBONO000683 [Email from Jane MacLeod to Rodric Williams and Andrew Parsons dated 20.03.19
1080. I WBON0000661 [Email from Tom Beezer to Jane MacLeod, David Cavender and Gideon Cohen dated
(14.03.19
1081. JPOL00167515 [Email exchange between Tom Beezer and Jane MacLeod dated 19.03.19- 20.03.19
1082. WWBONO000664 Email chain involving Amy Prime, Gideon Cohen, Tom Beezer, David Cavender and
Wane MacLeod dated 14.03.19- 15.03.19
1083. (WBON0001503 [Email from Gideon Cohen to Andrew Parsons, Tom Beezer, Dave Panaech, Amy
Prime, David Cavender and Owain Draper dated 18.03.19
1084. WBON0001504 Notes for recusal application - 18.03.19
1085. IPOL00364150 [Email chain involving Amy Prime, Gideon Cohen and Tom Beezer dated 19.03.19
1086. IPOL00364151 [Draft Witness Statement of Andrew Paul Parsons — 19.03.19
1087. WBONO000678 [Email chain involving Amy Prime, Gideon Cohen and Tom Beezer dated 19.03.19
1088. WBONO000674 Email chain involving David Cavender, Gideon Cohen, Andrew Parsons, Amy Prime,
ITom Beezer and Jane MacLeod dated 14.03.19— 17.03.19
1089. WBONO001512 [Email chain involving Stephanie Wood, Gideon Cohen and Tom Beezer dated 19.03.19
AC_207339548_5 52
WITN10390200
WITN10390200
No. I Document Reference Document
1090. WWBONO000680 Email chain involving Amy Prime, Gideon Cohen and Tom Beezer dated 19.03.19-—
20.03.19
1091. WBONO001514 Email chain involving Amy Prime, Gideon Cohen and Tom Beezer dated 19.03.19 —
20.03.19
1092. IWBONO001516 Email from Gideon Cohen to Tom Beezer, Andrew Parsons, Dave Panaech and Amy
Prime dated 20.03.19
1093. WBONO001519 Draft Witness Statement of Andrew Paul Parsons — 20.03.19
1094. WBONO001520 Email chain involving Andrew Parsons, Gideon Cohen, Amy Prime and Owain Draper
dated 20.03.19
1095. IPOL00023769 Email from Amy Prime to Jane MacLeod dated 20.03.19
1096. WBONO000685 Email chain involving Tom Beezer, Amy Prime and Jane MacLeod dated 20.03.19-
21.03.19
1097. WBONO000686 [Email chain involving Tom Beezer, Amy Prime and Jane MacLeod dated 20.03.19-
21.03.19
1098. I WBON0000687 Email chain involving Tom Beezer, Amy Prime and Jane MacLeod dated 20.03.19-
21.03.19
1099. IWBONO000684 [Email exchange between Andrew Parsons to Anthony de Garr Robinson dated 20.03.19
1100. WWBONO000688 Email from Amy Prime to David Cavender, Gideon Cohen, Stephanie Wood, Anthony
(Grabiner and Owain Draper dated 21.03.19
1101. IPOL00364171 [Email chain involving Gideon Cohen, Amy Prime, Stephanie Wood, Andrew Parsons
[Tom Beezer and Dave Panaech dated 22.03.19- 24.03.19
1102. JPOL00364172 Fifteenth Witness Statement of Andrew Paul Parsons — draft dated 24.03.19
1103. IPOL00364173 [Email chain involving Gideon Cohen, Amy Prime, Stephanie Wood, Andrew Parsons
[Tom Beezer and Dave Panaech dated 22.03.19— 25.03.19
1104. I WBONO000689 Email chain involving Gideon Cohen, Amy Prime, Stephanie Wood, Andrew Parsons
[Tom Beezer and Dave Panaech dated 22.03.19- 25.03.19
1105. IPOL00364175 Email chain involving Andrew Parsons, Gideon Cohen, Amy Prime, Stephanie Wood,
[Tom Beezer and Dave Panaech dated 22.03.19- 25.03.19
1106. IPOL00364177 [Email chain involving Andrew Parsons, Gideon Cohen, Amy Prime, Stephanie Wood,
[Tom Beezer and Dave Panaech dated 22.03.19- 25.03.19
1107. I WBONO000692 Email chain involving Amy Prime, Gideon Cohen, Stephanie Wood, Andrew Parsons,
\Tom Beezer, Dave Paenach, Owain Draper and Anthony Grabiner dated 22.03.19—
25.03.19
Email chain involving Amy Prime, Gideon Cohen, Stephanie Wood, Andrew Parsons,
1108. WBONOO00683 ITom Beezer, Dave Paenach, Owain Draper and Anthony Grabiner dated 22.03.19—
26.03.19
AC_207339548_5 53
WITN10390200
WITN10390200
No. I Document Reference Document
1109. POL00364183 Email chain involving Amy Prime, Gideon Cohen, Stephanie Wood, Andrew Parsons,
\Tom Beezer, Dave Paenach, Owain Draper and Anthony Grabiner dated 22.03.19—
26.03.19
Email chain involving Amy Prime, Gideon Cohen, Stephanie Wood, Andrew Parsons,
ITom Beezer, Dave Paenach, Owain Draper and Anthony Grabiner dated 22.03.19 —
26.03.19
1110. WBONO000694
Email chain involving Amy Prime, Gideon Cohen, Stephanie Wood, Andrew Parsons,
1111. WBON0001522 [Tom Beezer, Dave Paenach, Owain Draper and Anthony Grabiner dated 22.03.19 —
26.03.19
1112. POL00023950 Email from Amy Prime to Jane MacLeod and Rodric Williams dated 26.03.19
1113. WBONO001523 Email from Amy Prime to Gideon Cohen and Owain Draper dated 26.03.19
1114. WBONO001531 [Email exchange between Gideon Cohen, Amy Prime and Owain Draper dated 26.03.19
1115. POL00023239 Email chain involving Andrew Parsons, Rodric Williams, Amy Prime and Jane MacLeod
dated 26.03.19
1116. POL00269583 Letter from Freeths to Womble Bond Dickinson dated 26.03.19
1117. I WBONO000695 Email exchange between Gideon Cohen and Amy Prime dated 26.03.19
1118. IWBONOO000696 [Email chain involving Owain Draper, Amy Prime and Gideon Cohen dated 26.03.19
1119. POL00269584 Letter from Womble Bond Dickinson to Freeths dated 26.03.19
1120. WBONO001696 Letter from Freeths to Womble Bond Dickinson dated 27.03.19
1121. WWBONO000697 Email chain involving Owain Draper, Amy Prime, Gideon Cohen, David Cavender,
Anthony Grabiner, Miranda Bond, Dave Panaech, Charlie Temperley, Andrew Parsons,
Emma Campbell-Danesh, Victoria Brooks, Jonathan Gribben, Michael Wharton, Anna
Martin, Lucy Bremner, Beth Hooper, Katie Simmonds, Sushma MacGeoch, Rachel
Lawrie, Jane Atkinson and Mandy Robertson dated 27.03.19
1122. WBONOO000698 Email chain involving Owain Draper, Amy Prime, Gideon Cohen, David Cavender,
Anthony Grabiner, Miranda Bond, Dave Panaech, Charlie Temperley, Andrew Parsons,
[Emma Campbell-Danesh, Victoria Brooks, Jonathan Gribben, Michael Wharton, Anna
(Martin, Lucy Bremner, Beth Hooper, Katie Simmonds, Sushma MacGeoch, Rachel
Lawrie, Jane Atkinson and Mandy Robertson dated 27.03.19
1123. I WBONO000699 [Email chain involving Andrew Parsons, Owain Draper, Amy Prime, Gideon Cohen,
[David Cavender, Anthony Grabiner, Miranda Bond, Dave Panaech, Charlie Temperley,
[Emma Campbell-Danesh, Victoria Brooks, Jonathan Gribben, Michael Wharton, Anna
Martin, Lucy Bremner, Beth Hooper, Katie Simmonds, Sushma MacGeoch, Rachel
Lawrie, Jane Atkinson and Mandy Robertson dated 27.03.19
1124. I WBONO000700 [Email chain involving Owain Draper, Andrew Parsons, Angela Fraser, and Henry
farwick dated 27.03.19
1125. [POL00023208 Email chain involving Andrew Parsons, Jane MacLeod, David Neuberger, Gideon
(Cohen and Anthony Grabiner dated 12.04.19- 14.04.19
AC_207339548_5 54
WITN10390200
WITN10390200
No. I Document Reference Document
1126. I WBONO000706 Email chain involving Andrew Parsons, David Neuberger, Gideon Cohen and Anthony
(Grabiner dated 12.04.19- 14.04.19
1127. WBONO001572 Email chain involving Andrew Parsons, Jane MacLeod, David Neuberger and Anthony
(Grabiner dated 14.04.19 to 15.04.19
1128. WWBONO000707 Email chain involving Andrew Parsons, Anthony Grabiner, David Neuberger, Gideon
(Cohen and Owain Draper dated 12.04.19 — 15.04.19
1129. IPOL00006513 Email exchange between Andrew Parsons and Jane MacLeod dated 17.04.19
1130. WBONO000158 [Email from David Cavender to Tom Beezer, Andrew Parsons, Gideon Cohen and Owain
[Draper dated 21.03.19
1131. IWBONO000200 Email chain between Andrew Parsons, Tom Beezer, David Cavender, Gideon Cohen
jand Owain Draper dated 21.03.19 — 22.03.19
1132. IWBON0000323 [Email from Jane MacLeod to Andrew Parsons dated 05.04.19
1133. [POL00023941 [Email from Jane MacLeod to Andrew Parsons dated 08.04.19
1134. IPOL00270456 Email from David Cavender to Andrew Parsons and Amy Prime dated 10.04.19
1135. JPOL00270457 Draft Grounds of Appeal — 10.04.19
1136. WBONO001575 Email exchange between David Cavender and Andrew Parsons dated 16.04.19
1137. JPOL00023028 [Email from Andrew Parsons to Jane MacLeod and Rodric Williams dated 10.04.19
1138. JPOL00270458 [Draft Common Issues Judgment Appeal Advice dated 10.04.19
1139. IWBONO001576 Email from Andrew Parsons to David Cavender, Owain Draper and Gideon Cohen
dated 16.04.19
1140. JPOL00270870 Spreadsheet on whether POL should appeal or concede aspects of the Common Issues
judgment — 16.04.19
1141. JPOL00270936 Spreadsheet on whether POL should appeal or concede aspects of the Common Issues
Nudgment — 17.04.19
1142. WBONO001541 Email chain involving David Cavender, Andrew Parsons, and Rob Smith dated 09.04.19
1143. WBONO001571 [Email from Owain Draper to Andrew Parsons dated 12.04.19
1144. IWBONO0000291 Email from Anthony Grabiner to Andrew Parsons dated 08.05.19
1145. IPOL00284926 [Draft Grounds of Appeal — 08.05.19
1146. IPOL00006755 IMinutes of the Board Subcommittee meeting dated 24.04.19
1147. IPOL00023207 Lord Justice Coulson’s Order dated 09.05.19
1148. WBONO001717 [Email Amy Prime to Madeleine Collins dated 17.12.18
AC_207339548_5 55
WITN10390200
WITN10390200
No. I Document Reference Document
1149. POL00259980 Post Office's proposed approach to findings of fact following the Common Issues Trial
dated 17.12.18
1150. POL00269105 [Fourteenth Witness Statement of Andrew Paul Parsons dated 21.03.19
1151. WBONO000169 Email chain involving Anthony Grabiner, Andrew Parsons and Jane Macleod dated
(10.04.19
1152. I WBONO000172 [Email chain involving Anthony Grabiner, Andrew Parsons, Owain Draper, David
Neuberger and Gideon Cohen dated 12.04.19- 14.04.19
1153. IWWBONO000148 Email chain involving David Neuberger, Anthony Grabiner, David Cavender, Owain
Draper, Ben Foat, Alan Watts, Kirsten Massey, Tom Henderson, Amy Prime, Rodric
jilliams, Mark Underwood and Patrick Bourke dated 11.05.19 to 12.05.19
1154. IPOL00022933 Email exchange between John Grimmer and Amy Prime dated 18.03.19- 19.03.19
1155. (WBONO001697 Email from Brian Altman to Amy Prime dated 14.04.19
1156. IPOL00273923 Advice on the Common Issues Trial Judgment by Brian Altman dated 14.04.19
1157. IPOL00023223 Email chain involving Andrew Parsons, Rodric Williams, Brain Altman and Amy Prime
dated 14.04.19- 15.04.19
1158. WBONO001656 Email chain involving Lord William, Andrew Parson and Lee Bartlett dated 08.10.19
1159. WBONO001660 Email chain involving Lord William, Brain Altman, Andrew Parsons, Ben Foat, Rodric
illiams Charlyn Cruz, Catherin Emanual, Alan Watts, Ainsile Cranwell and Angela
Fraser dated 28.11.19
1160. WBONO000721 [Email chain involving Andrew Parsons, Jonathan Gribben, Catherine Emanuel, Alan
atts, Brain Altman and Lord William dated 08.12.19 - 09.12.19
1161. IWWBONO000723 Email from Jonathan Gribben to Anthony de Garr Robinson, Simon Henderson and
(Owain Draper dated 09.12.19
1162. WBONO000724 Email chain involving Anthony de Garr Robinson, Simon Henderson, Jonathan Gribben
jand Owain Draper dated 09.12.19
1163. WBONO001653 Letter from Womble Bond Dickinson to Freeths dated 25.09.19
1164. [POL00285257 Letter from Freeths to Womble Bond Dickinson dated 27.09.19
1165. WBON0000324 Draft letter from Womble Bond Dickinson to Freeths dated 30.09.19
1166. WBONO0000325 [Email from Lucy Bremner to Matthew Lenton dated 30.09.19
1167. IPOL00043025 Email chain involving Andrew Parsons, Matthew Lenton, Amy Prime and Lucy Bremner
Kdated 30.09.19 - 04.10.19
1168. I WBON0000137 Email chain involving Amy Prime, Andrew Parsons, Lucy Bremner, Jonathan Gribben
land Matthew Lenton dated 30.09.19- 01.10.19
1169. POL00285691 Letter from Womble Bond Dickinson to Freeths dated 03.10.19
AC_207339548_5 56
WITN10390200
WITN10390200
No. I Document Reference Document
1170. JPOL00285690 Letter from Womble Bond Dickinson to Freeths dated 03.10.19
1171. IPOL00285786 Fujitsu comments on relevant sections of EDQ — 06.12.17
1172. WBON0001430 Email chain involving Matthew Lenton, Amy Prime and Lucy Bremner dated 06.02.19-
(19.02.19
AC_207339548_5
57