WITNO9970100
WITNO9970100
Witness Name: Fintan Canavan
Statement No: WITN09970100
Dated: 22 August 2023
POST OFFICE HORIZON IT INQUIRY
First Witness Statement of Fintan Canavan
I Fintan Canavan will say as follows;
Introduction
1. I make this statement following receipt of a request by the Post Office Horizon
IT Inquiry under Rule 9 dated 31% July 2023 requesting information regarding
errors in disclosure provided to the Inquiry by Post Office Limited (POL).
2. I am a Partner in DAC Beachcroft solicitors but this statement is made in my
personal capacity and reflects my time seconded into POL as Inquiry Director
operating within the in-house Inquiry Team. DAC Beachcroft are not
instructed by POL in this Inquiry and, to the best of my knowledge, they have
not advised POL in regard to the Inquiry. I am aware they have provided
assistance to witnesses who have provided statements but I was not involved
in any of that work and have no information regarding the same.
3. I was assisted by Mr Nick Harvey a Partner in Peden & Reid solicitors in
Belfast in the preparation of this statement. I am grateful to him for his
assistance. The content is my own recollection.
Page 1 of 17
WITNO9970100
WITNO9970100
. Ijoined DAC Beachcroft in September 2020. In my earlier career I have
represented Core Participants in The Bloody Sunday Inquiry, The Inquiry into
Hyponatraemia Related Deaths in Children and the Historic Institutional
Abuse Inquiry. As a result of my experience I was contacted in or around July
2021 by a colleague in DAC Beachcroft and asked if I would be willing to be
seconded into POL for three (3) to four (4) months to assist with their internal
Inquiry Team.
. I confirmed I would be interested in that role but it was not until around
October 2021 that I was contacted again and met, via Teams, with
representatives from POL and was asked to join them. At that point the
Inquiry had previously been a non-statutory Inquiry but had progressed to the
present format although the full panorama of the Inquiry was not yet clear.
. There had been no announcements on the full anticipated scope and Phases
and it was anticipated, within POL, that the Inquiry process would last around
3 to 4 months only. I left the role as Inquiry Director in December 2022 and
have had no involvement with POL or the Inquiry process since then (save a
call around an issue about Confidentiality Undertakings) but have followed
some of the media coverage.
. Onjoining the Inquiry Team I reported directly to Benjamin Foat (Ben) the
General Counsel and the team I inherited consisted of a Senior Legal
Counsel, a small project team headed by Kevin Hutchinson and a Compliance
Director connected to the wider POL Legal and Governance unit headed by
Ben.
. Itwas immediately apparent that the scope of the Inquiry, the period under
Page 2 of 17
WITNO9970100
WITNO9970100
investigation and the issues involved meant the anticipated time scale and the
team within POL was inadequate. This created an immediate issue with the
budget provided for the project (a theme which underpinned much of the
work in the following 15 months).
9. Anumber of issues had to be addressed immediately. We needed to
increase the budget provision, increase the team and prepare for a longer
period of operation. The in house team grew to 6 Legal Counsel, one
paralegal with myself and a larger project management team led by Kevin.
10.As the various Phases of the Inquiry were set out I divided the internal team
into 3 smaller teams dealing with Phases 2 & 3, Phases 4 & 5, and Phases 6
& 7. There were 2 legal counsel in each team with the paralegal providing
floating support across the workstreams.
11. The role of the inhouse team was to assist in preparing the responses and
generating the knowledge base within POL and collating necessary material
as needed and directed by the external solicitors, Herbert Smith Freehills
(HSF). The POL team linked with the various business areas in POL to
identify issues and provide answers to the potential questions and to try to
front load the preparation to try and ensure we could respond to queries from
the Inquiry efficiently.
12.In my role I acted as a link between the GE and the team, the wider POL day
to day business departments (BAU), HSF, other external service providers
(e.g. Peters and Peters solicitors) as well as the lead in the team itself. I did
not act as a legal advisor nor did I provide legal advice. I was able to use my
previous inquiry experience to explain processes and clarification as to how
Page 3 of 17
WITNO9970100
WITNO9970100
the Inquiry would proceed and to try and give some idea as to what the
business would need to do to address the anticipated issues arising in later
Phases.
13.As the Inquiry developed and grew, the issues around the projected budget
and the need to secure a higher budget became the core focus and took up a
significant part of the time I spent in the role. A very significant amount of my
time was spent amending budgets, forecasting and projecting different
scenario budgets, seeking approvals for increased budgets and drawdown,
discussing ways to reduce fees with HSF and looking at ways to achieve the
same outcomes at a lower cost. This meant less and less of my time was
spent on the actual processes themselves.
14.1 attended the Freedom of Information Act/Subject Access Request
(FOIA/SAR) Steerco which assessed and made recommendations on
requests for information and data. I was not a quorum member of the
committee but sat on it in my capacity as Inquiry Director. This ensured that
there was a flow of information across the business about requested
disclosures or information recovered through the Inquiry process to cross
refer the issues to ensure POL remained compliant with it's FOIA duties but
also ensured that any information located which may be Inquiry relevant was
identified and referred to the relevant sub-team in POL and/or HSF.
15.1 spoke to Ben regularly and also spoke with Mr Nick Read, Mr Alistair
Cameron, Mr Richard Taylor and laterally Mr Jeff Smyth who were all GE
members. In addition to ad hoc conversations I had with them, these GE
members attended a regular Inquiry Steerco for updates on progress across
Page 4 of 17
WITNO9970100
WITNO9970100
many areas of the Inquiry and to give authority for steps needed. I believe the
full GE had effectively delegated day to day decisions on the Inquiry to the GE
members on Steerco to speed up the decision making process. The Inquiry
Team also prepared regular reports to the full GE and to the Board.
16.Before addressing the specific issues in the Request for Evidence I would like
to clarify the impression I had of those I dealt with in POL and their approach
to the Inquiry. The unambiguous approach within POL was a desire to co-
operate fully with the Inquiry, to discover what went wrong and to learn from
the past to improve the position. It was not disputed that things had gone
wrong.
17.Coming from a business position I think the legal process was not fully
understood but I believe all those involved in the Steerco and Inquiry
preparation were committed to that task. I had no doubt as to the sincere
desire to co-operate and had I not had that assurance I would not have
remained in the post I held.
Issues in the R9 request
18.POL disclosure came from two different streams; hard copy repositories, and
the electronic repositories.
19. The main electronic repository was held in a database managed by KPMG
referred to as Relativity. There were other smaller areas where electronic
data was stored (e.g. old laptops from departed staff and mimecast for the
email history).
Page 5 of 17
WITNO9970100
WITNO9970100
20.During my time steps were taken to amalgamate all electronic resources (e.g.
2
old emails on mimecast, older laptops etc.) onto the Relativity platform for
ease of access and a significant volume of older records from staff who had
left (prioritising those identified on witness lists supplied to the Inquiry) was
transferred over to that platform. From recollection I believe the Relativity
platform contained all the material from the earlier Horizon litigation and the
555 Group litigation. I believe there were in the region of of 50-60 million
documents on that platform.
. The process of transferring the email history from mimecast was essential as
the search facility on mimecast was not effective enough and, while it took
time, it was seen as an essential step. This is relevant to part of the
investigation in this hearing as it is a step taken (before any known issue)
taken by POL to improve their ability to respond and to provide the fullest
possible responses to any RQ request.
22. The issue of hard copy documents was a bigger issue. In considering that I
think the potted history (as I understood it) of POL is important.
23.POL in it's current iteration has developed from being a part of Royal Mail, a
sub division of Royal Mail and various iterations through that process. There
are large regional offices, smaller local offices, the main headquarters in
London and large offsite stores, like that in Winchester, as well as the older
historic records held in the Postal Museum (which also retains material related
to Royal Mail).
24. It was clear that the historic recording and storage of records and documents
was not ideal. As we identified any new hard copy repository we took steps to
Page 6 of 17
WITNO9970100
WITNO9970100
retrieve, catalogue and check all records looking particularly for records likely
to be relevant to issues in the Inquiry but especially for material which may
relate to earlier disclosures. [An example of this was an old store in
Londonderry Crown Office where records and 2 locked safes were located
during a tidying up process not related to the Inquiry. In addition to assessing
the records identified in the room steps were taken to locate a locksmith to
open both safes which revealed no relevant records as all material (to the
best of my recollection) pre-dated the Horizon system].
25.HSF were instructed by POL as the independent external legal advisor. This
instruction pre-dated my involvement and I understand it was influenced by
the involvement of HSF in earlier litigation and the initially anticipated short
duration of the non-statutory Inquiry. There was also a relationship with the
issues because HSF were involved in a redress process with POL connected
to the relevant issues.
26. The approach to each disclosure request (i.e. the search criteria, locations,
search terms etc.) was designed by HSF . Progress on the volume of
responsive documents and the likelihood of complying with R9 deadlines or
the need to seek extensions was communicated to POL through the Steerco
meetings. POL were not involved in the design of the search criteria but did
need to give authority for large budgetary spends (e.g. attendance at
Winchester by teams from HSF to conduct in-person searches of boxes and
records). POL instructed an experienced law firm and took their advice on the
process to be followed with HSF aware of, in agreement with and applying a
wide disclosure approach.
Page 7 of 17
WITNO9970100
WITNO9970100
27.As indicated earlier even the expanded Inquiry Team was not sufficiently large
to access the hard copy repositories or oversee the searches and returned
"hits" from the Relativity site nor to go out to the offsite storage and filter the
boxes which had any returns on search hits ( which numbered in the
thousands for each R9 request). The search criteria and search terms were
managed, designed and applied by HSF. I cannot recall specific discussions
but I do recall debate around these searches within POL and with HSF. The
concerns raised were to ensure the searches were sufficiently wide and not
overly restrictive.
28. There was a view expressed that the Inquiry should be invited to have
unrestricted access to the Relativity platform so that they could in effect
conduct their own searches for whatever documents they felt were relevant.
It was also felt that the search terms in each request should be set very widely
and that all returns be provided to the Inquiry. This was to be balanced as a
search against "Horizon" could produce numerous documents referring to
"horizon scanning", "looking over the horizon" etc. or references to
"prosecution/prosecuted" would return records for any and all prosecutions
regardless of their relevance to the issues in this Inquiry.
29. This approach reflected the view in POL that the Inquiry should be provided
with any material no matter how limited any relevance or potential relevance
was to a specific request. This was debated but there was a genuine concern
that this would have been seen as obstructive in a different way by trying to
hide the wood in the trees. I shared the concern that such an approach would
not have been welcome by the Inquiry despite the genuine bona fides of the
Page 8 of 17
WITNO9970100
WITNO9970100
offer.
30. The issues of "families" of documents, duplication of documents and so on
31.
were discussed in meetings in wide terms (i.e. there was no discussion about
any specific set of documents) but essentially HSF guided POL in the
disclosure while still operating under a general view that all relevant or
potentially relevant material should be disclosed.
It was noted that several copies of the same document could be sent to
several people in an email chain and that this was unlikely to be helpful to the
Inquiry providing the relevant emails were provided to ensure any comment in
emails on the documents were always provided. I recall some discussion on
ensuring that each version of a document was checked to ensure added
notes or comments (typed or handwritten) on a document were available but I
cannot locate any specific correspondence confirming that.
32. The drafting of the search terms and the identification of the relevant
documents was determined by HSF with a view that full co-operation was to
be provided. That approach was best seen in the hard copy document
searches.
33.When searches were conducted of the hard copy sites it was discussed and
agreed that even duplicates of documents should be recovered in a first level
review so that a second level review could look for hand written amendments
or comments not clearly visible on first examination which could be excluded if
simple duplication was a criteria for exclusion. However there remained an
unease that extreme volumes of disclosure could be portrayed as an attempt
to swamp the Inquiry and divert from the real content in the documents but
Page 9 of 17
WITNO9970100
WITNO9970100
that under disclosure could also be perceived as un-cooperative. The
preference was for over disclosure but within reason.
34. I recall (but not in specific detail) instances where documents were found in
searches which related to earlier R9 requests and these were voluntarily
disclosed to the Inquiry as soon as possible. Discussions took place (mainly
in Steerco but also between myself and HSF, Ben and others in the team)
about what had happened and how it could be avoided. It was a matter of
frustration to me personally but more so to Ben, Nick Read and Alistair
Cameron in the Steerco on each occasion more material was located.
35. The, admittedly late, but voluntary disclosure of records located during
searches reflects, in my view, the genuine desire to assist but also showed
the fractured and disjointed data storage in POL.
36.In response to the questions raised I can confirm and accept that the
provision of disclosure by POL was not ideal. The Disclosure Statements
provided by Ben did try to address those concerns but the internal processes
placed Ben, the team and HSF in a difficult position as new repositories would
be discovered as part of new processes devised to try and eradicate the drip
feed of small numbers of additional material being located at regular intervals.
37.1 believe there was an issue as to the discovery of a new repository on the
eve of a Disclosure Statement but I cannot recall which. This caused a
significant degree of anger in the GE and myself. I felt particularly bad for
Ben who was providing updates to the Inquiry based on the information
provided to him only to find that information was altered at the last minute.
These Disclosure Statements took a lot of time to prepare and placed a
Page 10 of 17
WITNO9970100
WITNO9970100
significant demand on Ben to complete them with HSF.
38.1 am aware of one specific disclosure issue covered in the press (racial
profiling material) and this occurred after my departure. I admit I was
surprised that this occurred in the way it did as I would have assumed there
was an Inquiry team member attending FOIA Steerco and so any relevant
material would have been brought to the attention of the Inquiry team
ensuring that the material was disclosed to the Inquiry before becoming
public. I do not understand how this material was released without the Inquiry
being informed.
39.1 have been provided with access to a Relativity site with access to my email
account and also some material from others in the POL Inquiry team from my
time in the role as Inquiry Director. Unfortunately even this moderated
material contains in the region of 55000 documents and so I have not been
able to isolate specific exchanges in regard to specific disclosure issues.
40.1 was not involved directly in the drafting of any of Ben's Disclosure
41.
statements but I was aware of the process. I am aware that Ben was
extremely frustrated that assurances provided within the business that hard
copy material had been located were subsequently overturned by further
material being located.
Material was located through a number of different events; a process to
access, review and index boxes held in Winchester (my memory is that there
were around 75,000 boxes stored there and that some did not have any form
of index or record as to their content) and this resulted in some R9 requests
being re-run against the new indices; spontaneous disclosures of material
Page 11 of 17
WITNO9970100
WITNO9970100
(e.g. laptops or files located when furniture was moved, offices checked or the
Londonderry Crown office discovery as a result of an unconnected process);
specific searches conducted at the request of the Inquiry team and directed
by the GE (a process was followed directing local managers to do a room by
room search which was clearly not fully followed as further disclosures arose
and more checks were needed).
42. Just before I left, and as a result of another discovery of material after the
directed searches of larger offices, it was decided that a full forensic search
be conducted and directed by an external agency with forensic search
experience. I was not involved in this as my workload was being downscaled
prior to my departure and I was concentrating solely on another series of
budget forecasting. Gemma Ludgate was responsible for overseeing that
particular process.
43.As a result of the significant volume of work involved in my role, my return to
my normal full caseload, no involvement with POL (save the media coverage
and a call about Confidentiality Undertakings) and the fact that I was
predominantly involved in budget management and prediction I cannot recall
many specific issues or disclosure matters.
44.1 do recall that disclosure, late disclosure, ad hoc disclosures and newly
discovered documents was a constant issue and this was raised regularly in
meetings and at Steerco. Each new incident was greeted with dismay and
taken very seriously.
45.1 do recall an issue arising in that a significant number of documents in
Relativity were identified but were password protected. This meant that even
Page 12 of 17
WITNO9970100
WITNO9970100
the Al search process nor a human review could not access the content and
so it was not possible to determine if the documents had any relevant material
included. It was discussed at Steerco and approval was given that we should
engage in a process of attempting to brute force these documents open using
a password cracking programme. As I understand it this involves a process of
"attacking" the document by bombarding it with passwords to see if it can be
forced open.
46. The discussion at Steerco (which included input from HSF) was that we
should try this and re-consider any material not opened and see if there were
any other options available at that point. I believe this process was still
underway when I left my role. This was a concern as we simply did not know
if there was any relevant material in those documents and so could not
precisely advise the Inquiry as to the position.
47.1 believe there are a number of different but interlinking reasons for the errors.
48. The first is the disjointed nature of POL itself. It is a massive business partly
State controlled and partly private sector.
49. Different areas of the business must operate within the budgets set for them
(largely imposed by Government but allocated with POL) and achieve their
BAU responsibilities and comply with the requirements of the Inquiry.
50. Rectification of legacy document storage does not feature highly in the
priorities of any area and there is a degree of passing the buck hoping other
areas will take up the responsibility. Again this goes to the budget allocations
and the need for those heading each business area or sub-area within those
workstreams did not allow for much leeway to allocate staff, time or money to
Page 13 of 17
WITNO9970100
WITNO9970100
document retention, identification and storage. I appreciate this may sound
critical of the business areas but it is not intended to be so. I always found the
BAU areas helpful but they were also mindful of the reality created by budget
restrictions.
51.Partly the numerous changes in "ownership" or structure means that there is
no single legacy management structure and even the records of what material
has been recorded can be disjointed.
52.Much of the material in Winchester has not been accurately indexed requiring
manual searches of high volumes of material. Boxes were found to contain
lottery receipts, memos, advertising brochures all in one space and if time and
money were available much of the material could be reviewed and destroyed
but that is not possible.
53. The Relativity platform is not easily understood by some of those in POL (I
include myself in that cohort) and the use of search criteria is also an art
rather than a science. HSF sought to identify relevant search criteria based
on each RQ request. Those terms were devised by and applied by HSF and
were not discussed with POL.
54.1 would have difficulty apportioning responsibility for any errors. Identifying
them as "errors" is, in my view, correct as I believe that each incident was
inadvertent rather than deliberate. I also do not believe that characterising
them as negligent would be correct. The issue is with the nature and number
of historic systems and the very tight budgets rather than as a failure to grasp
or address the issues. Nor is there any suggestion of a desire to avoid
dealing with or confronting the legacy matters which are the subject of this
Page 14 of 17
WITNO9970100
WITNO9970100
Inquiry.
55.In my opinion the wide ranging scope of POL's business is a significant
contributing factor to the problems.
56.1 also feel that the further a person is from the Head Office the less the
importance of material, no matter how limited, is appreciated. This was seen
when, after a directive that searches be conducted and assurances were
provided, a significant volume of material was located and a further Disclosure
Statement had to be provided to the Inquiry. This is not as a result of any
failing from the GE as regular information dissemination of POL updates by
Nick Read would regularly include reference to the Inquiry and the steps
needed to comply with POL's duties. It is also not a criticism of the individuals
who have to run post offices with all the duties and responsibilities that entails
and so taking time to conduct finger tip searches of every room, cupboard or
box is not your priority.
57.1 am aware that a process was underway between POL and KPMG with HSF
to work on the access to and responses from Relativity. Again at the point
when this was underway I was heavily involved in the re-forecasting of
budgets and cannot recall any specific detail about what was being arranged.
I cannot recall exactly what the issues were but I was copied in on some calls
and emails (I think to add seniority as it was felt this would push the process
faster).
58.1 do not recall seeing any response to the forensic search of hard copy sites
and it may be that the final report had not been complete. Gemma Ludgate
may have more detail on that or Diane Wills who adopted the majority of my
Page 15 of 17
WITNO9970100
WITNO9970100
role.
59.1 am not aware of what current processes and plans are now in place and the
change of legal provider may play a part in different approaches to search
criteria, review and disclosure.
60. In the period before I left I was aware that the issue was a concern and that it
was a matter of concern and there were steps in train to re-enforce the need
to address this.
61.1 do not believe it will be possible to ever assert that all issues are addressed.
The historic issues, human error in retention, indexing and storage coupled
with human error in review, identification and disclosure will be difficult to
avoid or eradicate and when the scope of the issues in this process are
considered there are potential pitfalls built in.
62.1 would say, in my view based on the experience I had with POL and the
people I met, and the approach of those in HSF responsible for applying the
disclosure criteria that any failure to disclose, late disclosure or other
disclosure issues are not caused or created as a deliberate attempt to
obstruct or thwart the Inquiry.
63.1 appreciate that any error is frustrating to the Inquiry and could be seen as an
attempt to obstruct when viewed by those who have suffered as a result of the
Horizon problems. I am sure that the team in place now are committed to the
best possible compliance.
I believe the content of this statement to be true.
Page 16 of 17
WITNO9970100
WITNO9970100
Dated: 22™4 August 2023
Page 17 of 17