19 January 2024 – Paul Patterson
Hide video Show video
(10.00 am)
Mr Beer: Good morning, sir, can you see and hear us?
Sir Wyn Williams: Yes, thank you very much.
Mr Beer: May I call Paul Patterson, please.
Sir Wyn Williams: Yes.
Paul Patterson
WILLIAM PAUL PATTERSON (sworn).
Questioned by Mr Beer
Mr Beer: Good morning, Mr Patterson. You know my name is Jason Beer and I ask questions on behalf of the Inquiry. Can you give us your full name, please?
Paul Patterson: Full name: William Paul Patterson.
Mr Beer: Thank you for coming to give evidence to the Inquiry today. You have been scheduled to give evidence today for many months now and today, I wish to be clear, is not a rerun of the evidence that you gave to the Business and Trade Committee on Tuesday of this week, not least because many of the issues addressed in your evidence to that committee will be addressed to you when you return to give evidence in Phases 5 and 6 of the Inquiry; do you understand?
Paul Patterson: Yes.
Mr Beer: In that connection, those watching the proceedings should understand that simply because an issue is not addressed by Mr Patterson today does not mean that it will not be addressed with him in due course. The Inquiry undertakes its work in a forensic manner, only asking questions of witnesses when it’s satisfied that all reasonably relevant material has been obtained so that questions can be put on a proper evidential foundation.
So you’re here today to give evidence primarily about the issues that you address in your third witness statement, which concerns ARQ data, you understand that?
Paul Patterson: Yes, I do.
Mr Beer: In particular, the reliability of that data, the use of it by Fujitsu and the Post Office, the provision of it to the Post Office in connection with criminal proceedings against subpostmasters; do you understand?
Paul Patterson: Yes.
Mr Beer: I am also going to ask you questions about matters in your second witness statement.
So can we start, please, by formally adducing the evidence in your second and third witness statements, not least so they can be made available to the public via the Inquiry’s website. Can we start with your second witness statement, please. You’ll find that in tab A2 of your hard copy bundle. The URN for it is FUJ00126035. It is dated 29 December 2022 and is 193 pages long, including its appendices. If you go to page 63, which is the last page of the statement itself, is that your signature?
Paul Patterson: Yes, it is.
Mr Beer: Are the contents of that witness statement true to the best of your knowledge and belief?
Paul Patterson: Yes, they are.
Mr Beer: Can we move to your third witness statement, please, which in your bundle is in tab A1. The URN for that is WITN06650300. That witness statement is dated 14 September 2023, it’s 103 pages long, including its appendices and, if you turn to page 80, you should find your signature –
Paul Patterson: Yes, it is.
Mr Beer: – is that your signature? Are the contents of that statement true to the best of your knowledge and belief?
Paul Patterson: Yes, they are.
Mr Beer: Thank you very much. Before I ask you questions of substance, this is the first occasion on which you’ve given oral evidence to the Inquiry and the first occasion on which a senior Fujitsu executive has given evidence to the Inquiry. I know in your first witness statement, which we published back on 2 December 2022, you gave an apology to subpostmasters and that you said something to equivalent effect on Tuesday of this week to Parliament.
Since then, Fujitsu, as a corporation, issued a public statement. Is there something you would like to say in that regard now?
Paul Patterson: Yes, Mr Beer. Thank you.
To the subpostmasters and their families, we apologise; Fujitsu apologises and is sorry for our part in this appalling miscarriage of justice. This Inquiry is examining those events forensically over many, many decades, which involved many parties, not least Fujitsu and the Post Office, but other organisations and individuals. We are determined to continue to support this Inquiry and get to the truth, wherever it lays, and, at the conclusion of the Inquiry and the guidance from this Inquiry, engage with Government on suitable contribution and redress to the subpostmasters and their families.
Thank you, Mr Beer.
Mr Beer: Thank you, can I start with your background, then, please. You are a Director of Fujitsu Services Limited, is that right –
Paul Patterson: Yes, it is.
Mr Beer: – which I’m going to call Fujitsu. I think your full title, at least according to the Internet and the Fujitsu webpage, is Corporate Executive Officer EVP; is that Executive Vice President?
Paul Patterson: Yes, it is.
Mr Beer: And the Co-CEO, is that Chief Executive Officer or Corporate Executive Officer?
Paul Patterson: It’s actually the first one, forgive me.
Mr Beer: So Chief Executive Officer –
Paul Patterson: Indeed, it is.
Mr Beer: – for the Europe region?
Paul Patterson: Indeed it is.
Mr Beer: So a Director, an Executive Vice President and Corporate Executive Officer for the Europe Region?
Paul Patterson: Yes.
Mr Beer: Thank you. When did you join Fujitsu?
Paul Patterson: In early 2010.
Mr Beer: What was your first role and job title?
Paul Patterson: I was hired into a sales role, a sales leadership role and I was the Sales Leader for the Private Sector.
Mr Beer: What did that job involve?
Paul Patterson: Very much focused on new business selling to Fujitsu’s customers in the private sector, as was then.
Mr Beer: What was your title again?
Paul Patterson: So Head of Sales for the Private Sector. We can – clearly we’ll confirm the individual titles correctly.
Mr Beer: Did that role have any connection to or responsibility for the Horizon IT System?
Paul Patterson: So subsequently my role changed and my responsibility changed. I was in a sales or sales leadership role for the UK private sector, then including the public sector, and then that evolved into being a European Head of Sales Role, as well. Throughout that – so in the early part – in the middle part of that, my responsibility also included the selling of the contract – new contracts to the Post Office.
Mr Beer: When was that?
Paul Patterson: So approximately 2012, I think ‘13, is when the Post Office new business selling responsibility came to – came into my responsibility.
Mr Beer: So before that, 2012, did your job bring you into connection with the Horizon IT system in any way?
Paul Patterson: I don’t believe so.
Mr Beer: After 2012, what was the extent of your connection to the Horizon IT System?
Paul Patterson: So my connection was to the Post Office. We had a number of contracts with the Post Office, one of them being Horizon, one of them being the telecoms contract, so their home phone and broadband contract, and it was my team who were responsible for the sale of the services associated with the new telecoms contract, and clearly any renewal or extension to the Horizon system subsequently.
Mr Beer: Did you remain in that role until you took up your present position in 2019?
Paul Patterson: No, I – that role changed, so I became a European Head of Sales and Business Development, and I did not have a direct team involved in the Post Office then.
Mr Beer: So how long did your involvement with the Post Office last?
Paul Patterson: I would – four years/five years.
Mr Beer: In that four to five-year period – let’s call it between 2012 and 2016/2017 – how regularly were you dealing with issues concerning the Horizon IT System?
Paul Patterson: So I wasn’t dealing with issues –
Mr Beer: Dealing with the Post Office Account, insofar as it concerned the Horizon IT System?
Paul Patterson: So my engagement and my sales team’s engagement was what was very much focused on the new business into the Post Office. It wasn’t dealing with the service delivery, albeit clearly at times in those conversations the Post Office may have asked me questions about service delivery and that will have been passed on to the relevant Service Delivery Team.
Mr Beer: That was my next question: to what extent were you cited on issues or problems with service delivery in that four to five-year period, so far as Horizon was concerned?
Paul Patterson: So I think, in several different meetings with Post Office executives when we were discussing the changes to the contract, so I think in about 2012/13, they were looking to change the contract structure into a tower structure. In some of those meetings we would have definitely made some reference to problems, given I was in the room, and that would have been handed back into the Service Delivery Team.
Mr Beer: To your recollection, did any of those engagements concern data reliability, data integrity or similar issues?
Paul Patterson: Well, so in the pack that I’ve got for – in this supplementary pack, I think, in today’s hearing, there is a reference to questions asked of me and a colleague by the CIO, which we passed on to, which was talking about that very point.
Mr Beer: I’m not going to go into that in detail today. I just want to get an overview today for the purposes of the questions I’m going to ask you subsequently of the extent to which you knew of issues concerning Horizon before you became CEO in 2019 or whether you were coming to these issues completely afresh in 2019. What would be the answer to that?
Paul Patterson: I – in 2019 in my appointment, I was, of course, aware that there were issues regarding prosecutions. Clearly there was the public case, as well, so I was aware on a personal level there were issues with the prosecutions. So I was aware of those topics.
Mr Beer: Before you became a CEO in 2019?
Paul Patterson: Yes.
Mr Beer: Okay. I think the email you’re referring to – I’m not going to go into it in detail today – shows that you had some involvement in and were briefed about the Second Sight investigations –
Paul Patterson: Yes.
Mr Beer: – and the impact of those on Fujitsu?
Paul Patterson: Yes.
Mr Beer: Thank you. So, overall, we shouldn’t get the impression, when we listen to your evidence in a moment, that you came to all of the issues concerning data integrity, the provision of information and evidence by Fujitsu to the Post Office and Post Office prosecutions afresh when you became CEO in 2019?
Paul Patterson: I joined – as we already discussed, I joined in 2010 and, certainly through my career inside the company latterly, I’ve become more and more aware of the issues and, clearly, during in this Inquiry, becoming more and more aware, at a very detailed level.
Mr Beer: Can I turn to your second witness statement, then, please. FUJ00126035, and it’ll come up on the screen for you. This is your second witness statement. You’ll see at the top it’s dated, as we’ve established already, 29 December 2022, so 13 months ago, or so.
If we look at paragraph 3, paragraphs 1 and 2 are usual introductions, but paragraph 3 at the foot of the page, as noted in your first corporate witness statement you do not have first-hand knowledge of many of the matters set out in the statement. You wish to reiterate at the outset how the information in the statement has been compiled. You have been assisted by a team of individuals within Fujitsu and Morrison Foerster. They’re the solicitors to Fujitsu for the purposes of this Inquiry; is that right?
Paul Patterson: Correct.
Mr Beer: “This is due to the vast amount of documentation and sources of evidence which have had to be reviewed for a time period stretching over 25 years. This team has provided to me the documents which were referenced in this statement and exhibited at …”
Then you give some reference numbers, and you exhibit 640 documents to this witness statement:
“… and which are the principal source of my knowledge of this statement’s contents.”
So the information in the statement that you’re giving is principally drawn from documents that have been provided to you by your team, essentially?
Paul Patterson: Correct.
Mr Beer: Then if we look at paragraph 4, please. You say:
“… responses to questions set out in the statement are generally drawn from documentary sources. These documents have been exhibited and/or referenced … The responses provided in this second … statement represent Fujitsu’s current understanding of the information available. Given that preparations for Phase 3 [were then] still ongoing, it may be that Fujitsu will need to supplement this corporate statement as further material is identified and made available to Core Participants.”
I’m going to skip over paragraphs 5 to 189 of this witness statement. That’s some 60 pages of the witness statement, which concerns Phase 3 issues in the Inquiry and about which we heard many weeks of evidence back at the end of 2022 and the beginning of 2023, so primary evidence from the witnesses concerned.
Can we pick up, please, what you say on page 61 of the witness statement, at the foot of the page, please, that is “Knowledge and Rectification of Bugs”. This section of the statement, indeed right to the end of the statement, addresses Fujitsu’s knowledge of and rectification of bugs in the Horizon system; is that right?
Paul Patterson: Correct.
Mr Beer: You say:
“As explained in Fujitsu’s opening statement … no complex IT system will ever be completely free of bugs, errors and defects (‘BEDs’). Fujitsu’s monitoring systems and provides seek to identify faults, log them as incidents, and then work to resolve them following the agreed incident management processes. Fujitsu also relies on incidents being reported by postmasters directly or by [Post Office Limited]. Many thousands of incidents have been logged since the inception of Horizon.”
Then over the page, please, or further down the page, paragraphs 191 and 192:
“In relation to the 29 [bugs, errors and defects] listed by Mr Justice Fraser in Appendix 1 to the Horizon Issues Judgment … the Inquiry has asked Fujitsu to provide details relating to the identification, investigation, communication and resolution of the [bugs, errors and defects].
“In February 2021, Fujitsu helped to prepare a report for [Post Office Limited] in relation to the 29 [bugs, errors and defects] identified by Mr Justice Fraser (the ‘BED [or bugs, errors and defects] Report’). This [bugs, errors and defects] Report has been disclosed to Core Participants and is exhibited [as your exhibit 260].”
So the sequence of events, just so we can get that clear and decode things, is Mr Justice Fraser produces his Horizon Issues judgment, that’s also known as Judgment Number 6., and that was in December 2019, correct?
Paul Patterson: Correct.
Mr Beer: 16 December 2019, to be precise. That contained a number of findings of fact, both in the body of the judgment and in an appendix to the judgment called Appendix 1, which you reference here, as to the existence of bugs, errors and defects in the Horizon system. That’s both, is this right, Legacy Horizon and Horizon Online?
Paul Patterson: I believe so.
Mr Beer: He analyses those 29 bugs, errors and defects in 418 pages of his judgment in Appendix 1 to his judgment, amounting to some 105 pages of closely typed text.
Then the next event is the event you refer to in February 2021, a report by Fujitsu to the Post Office. That’s 22 February 2021, so a year and three months after the judgment, Fujitsu writing a report to the Post Office. What do you understand the purpose of that report to have been?
Paul Patterson: I’m not quite sure I understand the question.
Mr Beer: So Mr Justice Fraser produces his judgment, finds the existence of 29 bugs, errors and defects, spends 105 pages analysing them, and then 13 months later Fujitsu write a report to the Post Office about those 29 bugs, errors and defects, and I was asking what was your understanding of the purpose of the writing of that report?
Paul Patterson: So in my – in the company’s second corporate statement we lay out details on the 29.
Mr Beer: Yes, I’m going to come to that in a moment.
Paul Patterson: I think – I don’t know, Mr Beer. I’m afraid I don’t know – I haven’t seen the physical report.
Mr Beer: It’s one of the exhibits to your statement, number 260 there. I’m not going to display it at the moment but I just want to understand, when the judge has found the existence of these 29 bugs, why a year and a month later Fujitsu is writing a report to the Post Office about those 29 bugs?
Paul Patterson: So I don’t know, Mr Beer.
Mr Beer: Okay. Let’s move on anyway, paragraph 193 of your witness statement. You say:
“In addition to the [Bugs, Errors and Defects] Report [the February 2021 report] Fujitsu has set out in Appendix 1 [that’s Appendix 1 to this very witness statement] a series of summaries addressing each of the 29 [bugs, errors and defects] and any sub-issues identified within those classifications. The [Bugs, Errors and Defects] Report and the summaries set out in Appendix 1 seek to build on the Technical Appendix [that’s Mr Justice Fraser] and have been prepared by a reference to a variety of sources. These summaries are indicative of, amongst other things: the investigation of each issue, the resolution of each issue, communication with other parties, including [Post Office Limited] and wider management, and the impact on branches.”
Then 194, please:
“The summaries in Appendix 1 are based on a review of contemporaneous documents primarily in the form of PinICLs, PEAKs and KELs, that have been identified as relevant to the relevant [bug, error or defect]. The summaries should be read in conjunction with these underlying records.”
Just to summarise, in Appendix 1 to this witness statement, you’ve set out 29 summaries relating to the 29 bugs, errors and defects found to have existed by Mr Justice Fraser, correct?
Paul Patterson: Correct.
Mr Beer: I’m going to look through some examples of those – I’m not going to go through all 29 – in a moment, so we can see what they look like and the kind of things they tell us. But all of this information, would you agree, was available to Fujitsu, indeed to you, because it’s in your witness statement, written in December 2022, but also earlier than that in the Bugs, Errors and Defects Report of February 2021?
Paul Patterson: Yes, because the information is there. That’s how we produced the report – we produced the documents in the second statement, correct.
Mr Beer: So in terms of something you said at the beginning of your evidence today about the Inquiry examining complex issues and Fujitsu wanting to wait until, essentially, the Inquiry has reported, in relation to this issue – and I’m not saying that for a moment that the Fujitsu summaries of the bugs, errors and defects are complete or should be taken to be the final word on each bug, error or defect – but, from Fujitsu’s perspective, is this right, as a company, for the last couple of years, it has known of the existence of these bugs, errors and defects at a corporate level?
Paul Patterson: Yes, in fact, all the bugs and errors have been known, at one level or not, for many, many years. Right from the very start of deployment of this system, there were bugs, errors and defects which were well known, to all parties, actually.
Mr Beer: You agree, I think, therefore, if we take the Fujitsu summaries, read together with the Bugs, Errors and Defects Report as a baseline, it follows that, at a senior level in the company, for the last couple of years, there has been corporate knowledge of the existence of these bugs?
Paul Patterson: I have known about these bugs because I’ve seen the report, yes, would be my answer.
Mr Beer: So there’s no need to wait for the conclusion of this Inquiry to find out at least this information because it’s already known to Fujitsu?
Paul Patterson: Yes, correct, and it’s in the statement. Correct.
Mr Beer: Can we go to the examples then, please. I’m going to look at four or five of the 29 bugs.
Can we start by looking at page 102 of the witness statement, please. Can we look at page 102 and 103 at the same time, please. That looks like that’s not going to be possible. I just wanted to get – ah, thank you.
This is what the appendices look like in to your witness statement. It lists the relevant bug, error or defect, this one is BED 2, the Callendar Square bug, then there are a series of bullet points, sometimes less than this, sometimes more than this, setting out in very summary form Fujitsu’s position on it; is that right?
Paul Patterson: Yes.
Mr Beer: The bullet point that always is first, says “Documents relating to Fujitsu’s knowledge, rectification and communication of the issue, including PinICLs PEAKs and KELs are set out in Appendix 2”, and then the relevant exhibit numbers are given. So it’s taking us, in the first bullet point off to the relevant exhibits and here there are about 25: 317 to 342?
Paul Patterson: Correct.
Mr Beer: I’m not, on each occasion, going to go back and look at the underlying material. If we can just read through, then the Callendar Square bug. You say:
“A problem existed in Horizon whereby, on occasion, a lock was not released, and a second process would then wait for a given time before it ‘timed out’ and reported an error saying it could not proceed. The problem could occur in various places in the Horizon applications. In the initial occurrences a reboot of the counter allowed the system to resume proper function with no data lost.”
Then the third bullet point:
“In some cases, in the Callendar Square branch in particular, the lock problem caused data to be lost when carrying out transfers between different ‘stock units’ thereby causing receipts and payments mismatches.”
So this is a bug, would you agree, that has a real impact on balancing because it causes receipts and payments mismatches?
Paul Patterson: Yes.
Mr Beer: Then the fourth bullet point:
“There appear to have been instances of these Riposte lock errors from at least September 2000 [so we get the start date]. In this instance, a postmaster reported an error … when trying to redeclare their cash. The call was discussed between various teams from [Post Office] and Fujitsu.”
Then you give some further examples of PEAKs.
Then a Known Error Log, fifth bullet point, and you give the reference:
“… advised that restarting Riposte, or rebooting the counter, would resolve incidents where a message reported a timeout waiting for a lock. Some of these also led to receipts and payments mismatches, which after investigation, were dealt with by the Fujitsu MSU team raising a BIMS report so [the Post Office] could issue an error notice (later known as a transaction correction) to the postmaster to allow them to reconcile the accounts. BIMS reports set out the progress to the resolution of a ‘business incident’. [The Post Office] would use the information from the BIMS report to carry out reconciliation or settlement …
“It was identified that an error in the underlying Escher Riposte software caused the lock problem. The issue was raised with Escher, who developed a fix. This was implemented in the S90 software release. S9 release was scheduled to start on 4 March 2006 for completion by 14 April 2006 … [By 22 March], the S90 migration report showed the counter release was 99.9% complete …
“Fujitsu monitored the issue … 27 March … an employee noted … that the timeout locks had ‘gone right down’ …”
So that, on the Fujitsu account of matters, was a problem that was first noticed in September 2000 and a fix was applied in March and April 2006, so five and a half years later. Yes?
Paul Patterson: Yes, that’s what it – yes.
Mr Beer: Now, in fact – this is an example where this shouldn’t be taken to be the last word on this – Mr Justice Fraser found that the bug continued in operation until 2010. But, on the Fujitsu account, this was a bug that was operative for close on a six-year period; is that right?
Paul Patterson: That’s what that says, yes.
Mr Beer: Let’s look at another example of a bug, please. Bug 3, the suspense account bug. That’s page 104 of your witness statement. So just page 104, please. Thank you. So bug, error or defect number 3, the suspense account bug. The usual opening paragraph. Bullet point 2:
“A change introduced into Horizon in July 2011 had the unintended consequence of leaving certain ‘orphaned’ records from November/December 2010 relating to a branch’s suspense account in a table in the branch database, rather than archiving them. The consequence was that, once a year, when an impacted branch produced its trading statement, if they had any amount in their suspense account, the suspense account records from 2010 were also pulled in so that the branch trading statement showed an erroneous amount in the suspense account.
“When the problem resurfaced a year later, a postmaster contacted Fujitsu and a PEAK … was raised on 25 February 2013 … Fujitsu then diagnosed the issue and identified 14 branches as being affected.
“… Fujitsu held a conference … with [the Post Office’s] Problem Manager … The orphaned records were subsequently removed by the Fujitsu development team. An extra set of checks were introduced in October 2013 so that if a similar problem surfaced in the future, an error message would be displayed to the postmaster telling them to contact the Horizon Service Desk.”
So here we can see we’re dealing with Horizon Online, yes –
Paul Patterson: Yes.
Mr Beer: – rather than Legacy Horizon, and it’s an issue that looks, on the Fujitsu account, to have lasted two years or so?
Paul Patterson: Yes.
Mr Beer: Can we turn, please, to a third bug, page 118 of your witness statement, please. This is the reversals bug. We can skip the first bullet point, which exhibits four documents. Second bullet point:
“A code fix distributed as part of the S30 release, caused a problem under certain circumstances due to faulty logic. On occasion, when a postmaster attempted to ‘reverse out’ a sum which had been ‘reminded in’ the balance showed double the initial amount rather than zero. According to the April 2003 Service Review Book … delivery of S30 commenced in April 2003 and by 2 May 2003, 2,135 banks were live.
“The initial issue … was reported by a postmaster on 24 April 2003 … sent to the Fujitsu [third line support, the Service Support Centre] on 28 April 2003, who identified that an error had occurred. A [Known Error Log] was raised … The issue was also routed to the Fujitsu MSU team so they could liaise with [the Post Office] who would then issue Transaction Corrections to rectify the accounts, following which the PEAK was to be routed back to Fujitsu so that Development could produce a code fix.”
Just stopping there, the bug was a doubling up problem, something that we’ve heard a lot about from subpostmasters themselves. We know from other documents, I’m not asking you to comment on this, that the postmaster that raised the issue that’s referred to in that third bullet point had raised the issue concerning £13,910. He had remmed in that sum as cash into his Horizon terminal but then, for a reason that’s not clear, needed to reverse that particular transaction out – might have got the wrong figure, person at the till said, “No, I don’t want you to put in that £13,910 cash” – whatever it was, wanted to reverse the transaction.
They did so and, instead of going back to zero, when they were remming out the transaction, the sum doubled to £27,830. Therefore, on the system, it showed that the subpostmaster should be holding cash of £27,820 relating to that transaction, whereas they wanted to show that they were holding none. Understood?
Paul Patterson: Yes.
Mr Beer: We can see that, from the next bullet point:
“On 30 April … the Fujitsu EPOSS Development team identified the coding error, and that it had been released with S30. An emergency fix was created … went live on 7 May 2003. Instructions for testers detailed how the fix was to be tested, to make sure that both the new problem and the original problem which S30 aimed to fix had been fixed.”
So this tells us – I think this is right, isn’t it – that the error was because of an attempted fix to another bug.
Paul Patterson: I think that is what it’s saying, yes.
Mr Beer: So if we look into the detail of it, we’ve got the underlying documents. The Horizon code – the problem was the person who’d written the Horizon code had applied the incorrect mathematical symbol to reverse the remming in, so, instead of applying the opposite mathematical symbol to what had been remmed in, it applied the same one as the operator had. So, instead of applying a minus following a plus, it applied a second plus, understand?
Paul Patterson: I understand the maths. I don’t know whether it was – all of that was behind this, but …
Mr Beer: Well, if we were to delve into the underlying material –
Paul Patterson: Okay.
Mr Beer: – that’s exactly what that would show.
Paul Patterson: Okay.
Mr Beer: So rather than I put some money in, plus; the reversal wants to reverse that, it should then apply a minus to take it back to zero. What it did was the mathematical symbol in the code was another plus sign, and so it doubled it.
Paul Patterson: Understood, and I agree.
Mr Beer: It was reported, according to that bullet point, in April and a fix to the problem caused by the earlier fix was applied in May 2003.
Can we look at a fourth bug, please, page 119. This is the data tree build failure, and this is split into a number of sub-issues. This is issue 1, which is the only one I’m going to address. Second bullet point:
“Fujitsu understands this issue was first reported to a branch on 10 November 1999 after a discrepancy in the accounts …
“A number of PinICLs for similar incidents were raised between February and May 2000 …
“A list of cross-domain problems was presented in the monthly Service Review Books to be discussed in the Service Review Forum with [Post Office]. The issue with the Dungannon branch was tracked in a number of these …
“To resolve the issue, Fujitsu implemented two changes – specific diagnostics to log a failure to build the data tree, and more error checking in the application code. [These] diagnostics were distributed to 99% of the estate by 16 May 2000 … this would allow recurrences of the problem to be monitored.”
Then the last bullet point: error checking was included in a later release.
So the issue was first reported on 10 November 1999, according to the second bullet point. Again, if we dug into the documents we would see that concerned a £43,000 discrepancy, so quite a sizeable sum of money, and the fix was rolled out, we see from this last bullet point, up until October 2000, yes?
Paul Patterson: Yes.
Mr Beer: Then, lastly, by way of example, page 146 of your witness statement, please, bug 18. Again, like the last bug that we looked at, there were a number of issues which have been split out, so sub-issues. This concerns the concurrent login bug and issue or sub-issue 2 of it. Second bullet point:
“Issue 2 concerned a receipts and payments mismatch that occurred due to a coding issue, which permitted a postmaster to transfer money from one counter to another while the first counter was being rolled over. The system should have prevented [this].
“The incident was raised with Fujitsu on 24 July 2000 … By [the 28th], Fujitsu had established what had happened after discussing the circumstances with the postmaster and confirmed [he] should not have been effectively able to looking into two counters with the same [ID] number. On 1 August 2000, a further incident … was recorded as another instance of the same issue …”
Next bullet point:
“A number of detailed analyses were carried out, with an interim diagnosis on 4 October … By 22 November … Fujitsu determined that this was a transient bug caused by two modules in the application using different methods to communicate with [each other]. … an earlier release (known as CI45) should have already fixed the issue.”
So a bug identified in July with a decision that an existing release by November 2000 should have fixed it. Okay, that can come down.
All of those 29 summaries are there for people to read, the Core Participants have got them and have had them for many months now.
Sir Wyn Williams: Before you carry on, Mr Beer, I just want to be clear about something.
In the previous examples you looked at with Mr Beer, Mr Patterson, you referenced documents which demonstrated that the Post Office was made aware of each of those bugs, more or less contemporaneously, all right? I’m not concerned about the detail of those documents.
In that last one that we looked at, unless I’ve missed it, you haven’t referenced a contemporaneous document or anything else to suggest that the Post Office was made aware of that bug. I just want to be clear about this: is your evidence that each and every bug was notified to the Post Office contemporaneously, or more or less contemporaneously, or is it the case that there may have been some bugs which were not?
Paul Patterson: My understanding –
Sir Wyn Williams: If I’ve missed the reference in BED 18, then please put me right, but I can’t see it at the moment.
Paul Patterson: And I can’t see it there either, Sir Wyn. So I think you are correct that the vast majority of bugs, errors and defects were shared, whether BED 18, in this example was, I would need to check.
Sir Wyn Williams: All right. Thank you.
Paul Patterson: I wouldn’t know off the top of my head.
Sir Wyn Williams: As Mr Beer has noted, you will be returning so there is no great urgency about it but I wanted to raise it while it was fresh in my mind. Thank you.
Mr Beer: Sir, you’re right to raise it. Generally, the summary where the Post Office has been notified, or information shared, says so.
So looking at the 29 as a whole, would you agree with the following points:
Firstly, in each case Fujitsu agrees that the bug, error or defect existed?
Paul Patterson: Yes.
Mr Beer: Secondly, on Fujitsu’s own assessment, by looking at this appendix, the bugs afflicted both Legacy Horizon and Horizon Online?
Paul Patterson: Yes.
Mr Beer: Thirdly, we can see by reference to this summary and, in particular, if we read it alongside the Bugs, Errors and Defects Report, which contains much more detail, we can see the date on which, according to Fujitsu at least, the bug, error or defect was recorded or recognised by Fujitsu?
Paul Patterson: Yes.
Mr Beer: Fourth, we can see the impact that Fujitsu assesses the bug to have that on the estate?
Paul Patterson: Yes.
Mr Beer: Fifthly, we can see, for most bugs, errors and defects, whether it was notified to the Post Office and, if so, when Fujitsu say the bug was communicated to the Post Office?
Paul Patterson: Yes.
Mr Beer: Sixthly, we can see that the earliest bug of the 29 bugs, errors and defects was in November 1999. It was one of the examples I took you to.
Paul Patterson: (The witness nodded)
Mr Beer: So that was in the course of the national rollout?
Paul Patterson: Yes, agreed.
Mr Beer: The latest, I’m not going to take you to it now, was May 2018, that was the bureau discrepancies bug, bug 14.
Paul Patterson: Agreed.
Mr Beer: So bugs, errors and defects afflicted the Horizon system, on Fujitsu’s own assessment, for a period of nearly two decades?
Paul Patterson: Yes.
Mr Beer: Then, lastly, we can see Fujitsu’s assessment of the length of time for which the bug was operative and, sometimes, that was a substantial period of time. The first one I took you to, Callendar Square, for at least six years or ten years by reference to Mr Justice Fraser’s findings?
Paul Patterson: Yes.
Mr Beer: So I think it follows from this that it’s plain that Fujitsu staff knew about bugs, errors and defects in Horizon, well before 2010?
Paul Patterson: Yes, I agree.
Mr Beer: Fujitsu staff knew of them, on Fujitsu’s own account, from at least November 1999?
Paul Patterson: Agreed.
Mr Beer: And that this recognition by Fujitsu, reflected in your witness statement here, doesn’t need any investigatory work to be undertaken by the Inquiry?
Paul Patterson: No, it doesn’t. It’s in the notes. I agree.
Mr Beer: It doesn’t need any application of judgement by anyone because it’s here in black and white, in Fujitsu’s own words, indeed in your own words?
Paul Patterson: Agreed.
Mr Beer: When did Fujitsu realise that the bugs, errors and defects in its Horizon system impacted on the evidence that was being relied on to investigate and prosecute subpostmasters for criminal offences?
Paul Patterson: Is that in my second statement?
Mr Beer: No. This going to your third statement, essentially, about Litigation Support.
Paul Patterson: So I don’t know the exact date off – just repeat the question again, Mr Beer. Sorry.
Mr Beer: Yes. When did Fujitsu realise that the bugs, errors and defects in its Horizon system impacted on the evidence that was being relied on to investigate and prosecute subpostmasters for criminal offences?
Paul Patterson: So I think in my – in the company’s second statement we draw attention to – we knew – the company knew several times that that evidence, that had been presented, needed to be corrected, given some bugs and errors, or the data needed to be rerun. So I think there will be several examples in the second statement which answers your question. I can’t give you the exact date on each and every one of them. I think in each particular ARQ request it would be applied differently.
Mr Beer: In other words, I’m asking when did Fujitsu put two and two together and realised they added up to four, four being “We need to tell the Post Office about these bugs, errors and defects, not because there’s a problem with the system that we’re selling to them, but because they’re prosecuting subpostmasters on the basis of the evidence that we’re providing to them”?
Paul Patterson: I think there’s lots of evidence of us informing the Post Office of that data that we’ve just discussed, bugs and errors, and how those bugs and errors did or did not impact the financial position as reported. What the Post Office did with that particular piece of data, Mr Beer, I do not believe Fujitsu knew at the time but, certainly latterly, of course the company became more aware that it was being used nearly solely for prosecutions.
Mr Beer: Would you agree that the 29 summaries that we’ve just looked at some examples of, revealing bugs, errors and defects in the Horizon System, ought to have been revealed to the Post Office for the purposes of its investigatory and prosecutorial functions?
Paul Patterson: So I don’t know if they were not.
Mr Beer: Yeah, that’s a different question, I’m asking would you agree that they ought to have been?
Paul Patterson: Oh, yes, I do.
Mr Beer: You know, I think, that Fujitsu employees provided witness statements to the Post Office for the purposes of the prosecution of subpostmasters and, speaking in general terms, these bugs, errors and defects did not find their way in a those witness statements. Do you know why?
Paul Patterson: I do not know why. I have seen examples of the witness statements and, on a personal level, I am surprised that that detail was not included in the witness statements given by Fujitsu staff to the Post Office, and I have seen some evidence of editing of witness statements by others.
Mr Beer: Where there was a proposal, I think you’re referring to, to include at least a reference to some of the bugs or some data integrity problems, and they were edited out?
Paul Patterson: Correct, Mr Beer.
Mr Beer: I have no doubt that you would regard that as shameful?
Paul Patterson: Yes, that would be one word I would use.
Mr Beer: What’s the other one?
Paul Patterson: Shameful, appalling. My understanding of how our laws work in this country, that all of the evidence should have been put in front of the subpostmaster, that the Post Office was relying on to prosecute them.
Mr Beer: Can we turn, please, to your third witness statement then, please. Now, the matters about which you speak in your third witness statement and we’re about to address through my questions to you, again, are generally the product of you having been provided with documents by your team or briefed by your team, in the same way as your second witness statement was created; is that right?
Paul Patterson: Correct.
Mr Beer: I’m going to ask you questions about the provision of Litigation Support by Fujitsu to the Post Office in connection with the Horizon system and, in particular, the use, the non-use, and the reliability of ARQ data Audit Record Query, or ARQ data.
Now, you start your third witness statement, perhaps naturally, with the contractual and other forms of formal documents that regulated or ought to have regulated the Litigation Support to be provided by Fujitsu to the Post Office. That’s where I’m going to start, so starting with the contract.
Are you aware that Fujitsu was contractually bound to provide evidence in support of Post Office prosecutions and civil proceedings?
Paul Patterson: Yes.
Mr Beer: Are you aware that Fujitsu operated a fraud and Litigation Support office?
Paul Patterson: Yes, I am aware.
Mr Beer: Which still exists, I think?
Paul Patterson: I don’t know if it does still exist, Mr Beer. I would need to check that.
Mr Beer: This office, the Fraud and Litigation Support Office, was to provide Horizon evidence to support prosecutions and civil actions, correct?
Paul Patterson: Yes, correct.
Mr Beer: Would you accept that, as Fujitsu was an integral part of the system supporting legal proceedings against subpostmasters, and knew that it was, it had a duty to ensure that the data that it supplied was accurate and complete?
Paul Patterson: Yes.
Mr Beer: Has what you discovered led you to the conclusion that the data supplied was not accurate and complete?
Paul Patterson: Yes, it has and we made – the company made that in this statement, actually, this number 3.
Mr Beer: Can we look at just what the contract said. So this is your third witness statement, WITN06650300, at page 3. Again, in general terms, I’m going to restrict my questions to what you’ve included in your witness statement, rather than looking at underlying materials, partly because of the limitations of your evidence, given your position, partly because I suspect it’ll turn into an exercise of you saying you weren’t in the relevant post at the time and you didn’t see this document or that document at the relevant time, but you can read the document like the rest of us, and I don’t want you just interpreting documents.
Can we look at paragraphs 6 and 7 then, please. You say:
“from the outset of Horizon, Fujitsu has been required by contract to maintain an audit trail of ‘all transactions and events …’ (see for example paragraph 3.1 of Schedule A03 to the Codified Agreement … of 28 July 1999 … and paragraph 3.1 of Schedule D5 to version 13 of the Codified Agreement of 23 November 2020 …”
I’m not going to ask you any questions about the agreement of 2020 because prosecutions had stopped by then.
“This contractual obligation flowed from requirement 699 contained at paragraph 1.102 of Schedule A15 to the 1999 Codified Agreement. In particular, requirement 699 notes at:
“… 1.102.6: ‘The content of the audit trail should be agreed with [Post Office Counters Limited] by a date consistent with the Project Plan’.
“… 1.102.9: ‘The audit trail shall have a level of security such that it cannot be altered or deleted.’.”
No need to read 102.11.
So you focused in these two paragraphs on the obligations of Fujitsu arising from requirement 699 of the Codified Agreement?
Paul Patterson: Correct.
Mr Beer: They are all about the duty to maintain and provide what’s called an audit trail?
Paul Patterson: Correct.
Mr Beer: Were you told about, and did you take into account in what you said in this witness statement, a separate requirement in the Codified Agreement concerning what was described as prosecution support, that’s requirement 829?
Paul Patterson: Do I –
Mr Beer: No.
Paul Patterson: I think we would have done, in our response to – in our evidence in here in number 3. Yes, we would have done.
Mr Beer: You see these paragraphs – there’s a series of requirements in the contract –
Paul Patterson: Yes.
Mr Beer: – and you’re focusing all about 699, which is about audit, either operational audit or commercial audit by Auditors. There’s a separate series of requirements concerning the provision of evident for the purposes of prosecution, which spring up from requirement 829, which you don’t analyse here at all. Did you know about that separate requirement, the 829 requirement?
Paul Patterson: So I am aware of the – I didn’t know the number but I was aware, and the company was aware, that there was a prosecutionary (sic) support obligation in the contract, yes.
Mr Beer: Okay. In any event, let’s look at what you do talk about concerning the audit requirement, which may be a very different thing to prosecution support.
Paul Patterson: Okay.
Mr Beer: In that paragraph at the bottom of the page we’ve got there, you remind us that the Codified Agreement said:
“The audit trail shall have a level of security such that it cannot be altered or deleted.”
Is it now recognised and accepted by Fujitsu that Fujitsu could and did make insertions and amendments into data which had an impact on branch accounts?
Paul Patterson: The way I believe the system worked for discovering – for the audit trail was to take the raw data and take a copy of that to then provide the ARQs. So I think, in the ARQ data, you could filter out or add data to that. So yes.
Mr Beer: Do you accept this meant that Fujitsu could and did alter the audit trail data?
Paul Patterson: I think it does mean that, yes.
Mr Beer: Ie do the opposite to what the requirement in 102.9 says?
Paul Patterson: I think the changes or any adjustments were agreed with the Post Office before any action would or would not be – hence the bugs, errors and defects. So I don’t think it was a secret intervention. I think it was discussed, you know, “This bug, this error, causes this: make change”.
Mr Beer: But I think you would accept that audit data should have been an exact reflection of the transactions taking place at the branch, no more and no less?
Paul Patterson: I do agree with that, and I think the underlying data in the message store was exactly that.
Mr Beer: That wasn’t what was given in the audit data?
Paul Patterson: No.
Mr Beer: Can we turn to page 8, please, where you set out for us a flowchart, at the top of the page, thank you, taken from a prosecution support process document of 29 February 2005, so a Fujitsu policy document. I just want to look at the policy document. This is one of the rare occasions when I’m going to delve into the underlying material that’s an annex to your witness statement. It’s just so we can understand some foundational terms, get those locked down for our later discussion.
The underlying document from which this diagram is taken is FUJ00152209.
This is one of the exhibits to your witness statement, the 11th exhibit to your witness statement, and we can see the date of it, as I’ve just said, in the top right there, 29 February 2005. We can see that it’s version 2, from the top. If we just go over the page, please. We can see from that table at the top there, I think the fourth entry on that table, that version 1 of this document was dated 26 November 2002.
Paul Patterson: Yes.
Mr Beer: Let’s work from this version, the one that was operative from 29 February 2005. I’m afraid we’re going to have to go through a bit of it, just to understand some terminology and the process that was intended, to start with. Can we start, please, with page 8, “Scope” of document, if we can just look at the top half of the page:
“This document sets out the procedures to be adopted by Post Office Account’s Prosecution Support Service for managing and dealing with Audit Record Queries for investigation and support purposes including the:
“Undertaking of Audit Record Queries;
“Presentation of transaction records extracted by Audit Record Queries;
“Analysis of appropriate records and logs;
“Preparation of witness statements of fact in relation to Audit Record Queries;
“Attendance at court by relevant employees to give evidence in respect of witness statements;
“Undertaking of additional litigation/prosecution support activities as may be requested on a case-by-case basis on the instruction of legal counsel.”
I’m going to skip the next paragraph and then on to:
“ARQs in support of potential prosecution will be obtained solely from the Horizon System Audit Archive/Server. The method by which the text of this data is protected is described in the Audit Trail Functional Specification, evidence in support of data integrity will be sourced from the Audit Archive/Server and Post Office Account business logs. All access to audit data is restricted to named individuals via dedicated workstations located in a secure environment. This is consistent with the security controls employed for the existing service. Supporting evidence is sourced from relevant business records and logs.”
Two types of requests:
“Audit Record Query only [involving] the extraction [of] audit archive of records … for an outlet.
“[ARQ] plus witness statement … involves the extraction [of] the audit archive of records … plus the provision of a [particular] witness statement …”
Yes?
Paul Patterson: Yes.
Mr Beer: Then if we go over to page 9, please. There’s something about the history in the penultimate paragraph on this page:
“The provision of prosecution support (specifically the provision of witness statements of fact) was … not formalised and was provided on a ‘without prejudice subject to contract’ basis pending the receipt of [a] Change Request. Prosecution support for the existing system is now provided as part of the Prosecution Support [System]. This document outlines the operational approach to this service.”
So that’s a reflection, is it, that, before this document, there wasn’t a formalised statement or policy on the provision of prosecution support?
Paul Patterson: That is my understanding.
Mr Beer: Then if we can go forwards, please, to page 10. If we read paragraph 3.2 at the bottom. There is provision here about the limits or the limitations on ARQs:
“The number of [ARQs] requested by Post Office in accordance with investigation or prosecution shall be the first to be met per year of [720 or 15,000 query days] on a rolling basis.
“With no more than the first to be met in any calendar month of
“60 queries … or
“1,250 query days.”
“Any [ARQs] over and above the 720 maximum will be rolled over to the next 12-month period and count towards the total for the next year.
“Post Office may vary the aggregate limits of [ARQs] [between the limits set out above and] the following substitutes for those limits …”
There are some different figures required or provided for.
Then if we go to the foot of this page, thank you:
“Each [Audit Record Query] shall relate only to an individual Outlet.
“[ARQs] are limited to specific types of information/data fields. These are”, then they are listed.
Then page 14, please, under the heading “Prosecution Support”:
“In addition to the details at 3.4 above [which we’ve just looked at] Post Office shall wherever possible, advise on the relevant section of the [ARQ] Form whether an associated witness statement is required (see Appendix 1)”, which we’ll look at in a moment.
Then “Scope”:
“Post Office Account [that means Fujitsu] shall, in relation to an [ARQ], at the request of Post Office:
“[1] Analyse appropriate Horizon Helpdesk and non-polling reports for the specific search criteria … in order to check the integrity of transactions extracted …
“[2] Analyse fault logs for the devices from which the records of transactions were obtained to check the integrity of transactions;
“[3] Provide witness statements of fact in relation to that [ARQ];
“The above analyses and witness statements will be undertaken in respect of a maximum of 250 [ARQs] per year …”
Fifth:
“[Fujitsu will] Provide for the attendance at court by the person who has provided a witness statement as identified above to give evidence in support of that witness statement … a maximum of 100 days [a] year.”
Then at page 19, please, we see, at the top of the page there, under paragraph 7.0, the diagram which you have cut into your witness statement –
Paul Patterson: Yes.
Mr Beer: – which we’ll come back to in a moment.
Then, at the foot of the page, 7.1 and 7.1.1, there’s a list of the nine steps to be taken when an ARQ request is received. So 7.1.1, team member shall “identify the search criteria”.
Then if we just go over the page, 2, they shall create an audit trail of the request.
Then 3, search the files required to complete the report.
4, select and retrieve the files.
5, generate the message store.
6, use a tool called RQuery to select the files per the search criteria.
Over the page, please, 7, burn the data onto a closed CD, along with a Word document with an explanation.
8: carry out a virus check.
9: dispatch it.
So the nine steps in the process are described.
Then page 21, please. You’ll remember that there were four other things under prosecution support that Fujitsu could do, and this lists them out. If we look at 7.2.1, so the first of them, 2.1, is “Check Horizon System Helpdesk Logs”:
“Problems or faults at a post office outlet logged with the [HSH] will be examined using the search criteria …”
So this is, if the subpostmaster has called an issue in, it ought to be discovered by searching the Helpdesk logs, yes?
Then, secondly, over the page, the second thing as well, as the production of ARQ data, that was to be done, was an analysis of non-polling reports:
“Non-polling reports shall be reviewed for the outlet in question …”
Do you know what non-polling is?
Paul Patterson: I’m assuming it’s about the network and connecting to the main database and –
Mr Beer: Broadly, yes. So conduct an analysis of non-polling reports.
Paul Patterson: Mm-hm.
Mr Beer: The third thing is analyse the fault logs, so any relevant PinICLs in PowerHelp logs will be reviewed through the PEAK system to identify any recorded faults that might affect the integrity or admissibility of the audit archive from which the ARQ queries are extracted:
“The PEAK log will detail the error relating to the site equipment [and] service in question.”
Then the fourth add-on, “Complete a witness statement of fact”:
“[Prosecution support] will provide a witness statement of fact … as far as possible to be undertaken by the person responsible for the actioning of the work … so as to retain continuity of evidence …”
Then 7.4.1, about witness statements of fact:
“Any material or otherwise pertinent information shall be recorded and included in the relevant witness statement of fact.
“Requirements for witness statements … shall be completed by the did from [Prosecution Support] who completed the request.
“The statement shall follow the standard format and layout for witness statements of fact provided in evidence. Contents of witness statements of fact are flexible depending on specific requirements … and the knowledge of the witness giving the statement. An example of a witness statement of fact is provided in Appendix 2.”
Let’s just go and look at that, please. That’s page 29, I think, of this document. So there’s a template or boilerplate witness statement. If you just look at paragraphs 1 and 2 and just read those to yourself.
Paul Patterson: Yes.
Mr Beer: Then over the page, you see there’s an explanation, in C, of the system, and in D, and in E. Then if we go forwards to page 32, please, the foot of the page of 32:
“During audit extractions the following controls apply …”
Then they are listed out between 1 and 10, if you just keep scrolling, thank you.
Then there’s some deletions and then this:
“ARQ [whatever the number was] was received on [whatever the date was] and asked for information in connection with the Post Office at [whatever the FAD code of the post office was]. I produce a copy of ARQ … as [an exhibit number]. On various dates and at various times between [two dates], I undertook extractions of data held on the Horizon system in accordance with the requirements of [something] and followed the procedure outlined above. I produced the resultant CD as [X, as an exhibit number].”
That’s all we need to look at there. If we go back, then, please, to page 22 of the prosecution support document. We were looking at witness statements of fact in 7.2.4.1 at the foot, and we’d reached halfway through paragraph 3:
“For each request, Post Office Limited and [prosecution support] will agree relevant matters (such as those listed below) which should be covered in the witness statement of fact (based on the knowledge of the witness):
“[1] Identification information about the author …
“A summary of the previous manual system used by the Post Office before Horizon.
“A summary of Horizon and what information is recorded.
“How consistent time is recorded within [Horizon].
“The types of report that can be generated on a counter by a clerk.”
Over the page:
“The transfer of accounts from Post Office main accounts department.
“A brief overview of all applications …
“How data is passed from the counter to [the archive].
“The process for extracting information for [ARQs] and the controls in place to protect and ensure the integrity of that data.
“An analysis of the [ARQ], when the [ARQ] form was received and the dates when the audit data extraction took place …
“A summary of the evidence provided for the request.”
Then 7.2.4.2, “Court attendance in support of Witness Statement of Fact”:
“The author of a witness statement of fact may be required to attend court in order to bear testimony to the facts.”
2.5, “Provision of exhibits”. This will generally comprise one of the following four: CDs, which we’ve seen; HSH logs, which we’ve seen reference to; non-polling reports, which we’ve seen reference to; and fault logs. They are back references, essentially, to paragraphs 7.2.1, 2 and 3 of the document.
Then over the page, please, sorry, to page 25, under the heading “Additional Prosecution Support”:
“There may be occasions when information is requested which exceeds that provided for …”
Under 8.2, “Expert witness” evidence:
“To offer all the available evidence without it being requested will only serve to flood the courtroom with documentation. For this reason, expert in-depth analysis and detailed ‘expert’ witness statements (as opposed to witness statements of fact) are rarely required.
“It is … conceivable that, given the size and complexity of the Horizon system, the integrity of the witness statements of fact may be challenged by defence council in order to discredit a prosecution. In these cases additional, granular detail about the technical working and integrity of various systems … may be required if only for ‘unused material’.”
Then there’s a list of the types of expert evidence that could be called upon to be provided. Above that:
“Expert witnesses could comprise anyone within the Post Office Account or its approved contractors who would be called upon to provide and testify to this additional evidence.”
Right, we can stop there, looking at that policy document.
Would you agree that this provides quite comprehensive guidance on the provision of prosecution support by the Fujitsu Prosecution Support Service to the Post Office?
Paul Patterson: Yes, it does.
Mr Beer: Would you agree that it recognises, on its face, a difference between evidence of fact and expert evidence –
Paul Patterson: Yes.
Mr Beer: – and that it treats them differently?
Paul Patterson: Yes.
Mr Beer: Would you agree it sets out the steps to be taken in each case to obtain and then to disclose ARQ data –
Paul Patterson: Yes.
Mr Beer: – and it reflects those steps or requires those steps to be reflected and spoken to in a witness statement?
Paul Patterson: And more. So I think the ARQ data alone is not enough and, in our corporate statement, we say that also.
Mr Beer: Yes. We’re going to come to that probably after the break, the important point you make in paragraph 19 of your statement –
Paul Patterson: Yes.
Mr Beer: – that ARQ was never enough.
Paul Patterson: Yes, and I think that document shows that there is a range of information that the subpostmaster should have been presented with.
Mr Beer: If we go back to the diagram on page 8 of your third witness statement, please, WITN06650300, page 8, please. If we look at the diagram at the top, you’ll see that it splits immediately – if that can be blown up, the diagram, please, thank you – between, on the right-hand side, an ARQ form which is for prosecution support and, on the left-hand side, seemingly one which is not. Yes?
Paul Patterson: Yes.
Mr Beer: It treats them differently. You’ll see that, on the right-hand side, as we’ve just seen in the policy document, that step 1 includes checking Helpdesk logs, we’ve seen that; the second step is to analyse non-polling reports; and the third step is analyse the PEAKs and, as we’ve seen in the policy document, that’s all about integrity of data.
Paul Patterson: Yes.
Mr Beer: It doesn’t include checking the Known Error Log – that’s neither in the diagram nor in the policy – does it?
Paul Patterson: No, it doesn’t.
Mr Beer: So it’s not in the diagram, it’s not in the policy and, if we looked, it’s not in the witness statement either, the boilerplate witness statement. Do you know why that is, that if you’re wanting to look at the integrity of Horizon data, one wouldn’t look at the Known Error Log?
Paul Patterson: I don’t know why they wouldn’t have done and I would have expected a more holistic assessment of the entire environment that a subpostmaster was using, and so I would have expected error logs and other matters to be presented and considered.
Mr Beer: In your reading of the materials, in your investigation of the issues and in the briefings you have received, did you notice any reluctance on the part of Fujitsu in the past to reveal the existence of a thing called the Known Error Log?
Paul Patterson: There is, in the submission to the Inquiry today for number 3, there is evidence of that, where “Don’t share with the Post Office yet”. I don’t know the individual situation, whether was subsequently shared with the Post Office but there was certainly those reluctance. Whether that was just for completeness, completeness to make sure had what we shared with the Post Office was complete versus, I think – it may well be there’s definitely evidence in submissions from – in this submission, around exactly what you just described.
Mr Beer: Do you know why Fujitsu might be reluctant to reveal even the existence of something called the Known Error Log?
Paul Patterson: No, I – it – the title is “Known Error Log”. It’s not unusual in a large system of – certainly of this size, that there will be errors and known errors and, certainly, from the very outset, there were lists of them and communication between all parties. How that was communicated to subpostmasters, I think, is slightly different but known errors were known, and lots of people knew them. With a particular one, Mr Beer, to your question earlier, that might be a timing thing versus not trying to share it.
Mr Beer: I’m not, at the moment, delving into any individual cases as to why the Known Error Log was not revealed to a subpostmaster in a prosecution; I’m asking why it’s missing from the process.
Paul Patterson: I have no idea why it’s not.
Mr Beer: Sir, that would be an appropriate moment, if it’s convenient to you, to take the morning break until 11.45.
Sir Wyn Williams: That’s fine by me, so we’ll reconvene at 11.45.
Mr Beer: Thank you very much, sir.
(11.29 am)
(A short break)
(11.45 am)
Mr Beer: Good morning, sir, can you still see and hear us?
Sir Wyn Williams: Yes, thank you.
Mr Beer: Thank you.
Mr Patterson, we were just dealing with the diagram which reduces to diagrammatic form the main parts of the process that we saw in the 2005 process document. I think you’ll agree that it – either the process or the diagram – does not include, as part of the process, checking event logs –
Paul Patterson: Yes.
Mr Beer: – including the NT event log?
Paul Patterson: Mm-hm.
Mr Beer: Yes?
Paul Patterson: Yes.
Mr Beer: I think elsewhere in your statement – I’m not going to take you to them, I think paragraph 89 and 100 in your witness statement – you speak to the fact that some errors, which were not otherwise picked up, were recorded in such error logs?
Paul Patterson: Yes.
Mr Beer: You tell us in your statement that checking event logs was only considered as part of the routine process to be undertaken, after the locking problems were discovered in 2008. Do you know why checking event logs was only considered as part of the routine process after 2008?
Paul Patterson: No, I do not, Mr Beer.
Mr Beer: When you describe that for us in your witness statement, you say that you understood or it is understood to have become part of the process in 2008. How do you know that it was or it is understood to have become part of the process after 2008?
Paul Patterson: Only from the information that we’ve gathered from conversations and documentation from investigations around the process. That is what I think I’ve reflected in the submission.
Mr Beer: The process that we see in the diagram and in the policy did not include checking the message store for any notes, for example, left by SSC staff, correct?
Paul Patterson: Correct.
Mr Beer: Therefore, if SSC staff left messages when they had inserted data into branch accounts – we’ve heard about this from Anne Chambers – that would not be revealed by the process undertaken?
Paul Patterson: No, it wouldn’t be.
Mr Beer: Do you know why that check was not built in to the process: checking the message store for notes left by SSC staff –
Paul Patterson: No, I do not.
Mr Beer: – that might record or reflect the fact that they had inserted data into branch accounts?
Paul Patterson: So I do not, Mr Beer, why.
Mr Beer: As a result of those things that were not done, checking the Known Error Logs, not looking at event logs, including the NT event log, not checking the message store for notes left by SSC staff, means that, in the data gathering process, when the Post Office made a request for ARQ data, Fujitsu did not provide everything that was required to reveal whether Horizon was working properly at a particular branch at a particular time; do you agree?
Paul Patterson: I think the document requesting the ARQ – I think you said we saw earlier about witness statement or no witness statement – that document laid out the criteria for the search.
Mr Beer: Yes.
Paul Patterson: It did not include those points of –
Mr Beer: Those three things?
Paul Patterson: Yes, those three things that we’ve just agreed on, sir, and I don’t know why.
Mr Beer: Would you agree that the failure to include, whether as part of the process or a witness statement that reflected the steps that were taken as part of that process, mean that Fujitsu did not provide everything that was required to reveal whether Horizon was working at a particular branch at a particular time properly?
Paul Patterson: Yes, I think the witness statement and other evidence should have been far more comprehensive before it was placed in front of a subpostmaster. Yes.
Mr Beer: Not just a witness statement, the steps that were taken?
Paul Patterson: Oh, no, indeed, yes, yes.
Mr Beer: Because the witness statement should only be a reflection of what has been done.
Paul Patterson: Agreed, agreed.
Mr Beer: Additionally we’ve heard evidence, this week in fact, that members of third line support, the SSC, undertook a process of filtering ARQ data before it was provided to the Post Office, and that filtering of data meant that some relevant data may not have been provided to the Post Office. That part of the process, the SSC getting involved and filtering data out, is not described in either the policy, the diagram or the witness statement, is it?
Paul Patterson: No, it’s not.
Mr Beer: So it’s not in the process map we see here?
Paul Patterson: No, it’s not.
Mr Beer: It’s not in the broader policy or, indeed, in any other document that describes the full process, and it’s not in the boilerplate witness statement?
Paul Patterson: No, it is not.
Mr Beer: Indeed, would you agree the witness statement, the boilerplate witness statement, gives the impression that all of the raw data that has been obtained within the relevant date ranges has been extracted and provided to the Post Office?
Paul Patterson: Yes, it does.
Mr Beer: Whereas, in fact, there’s a step in the process that has not been revealed to the subpostmaster or to the court?
Paul Patterson: Agreed.
Mr Beer: So, if the evidence that we’ve heard from Fujitsu witnesses this week is correct, then a witness statement that followed the template and didn’t mention the filtering out exercise would mean that the witness statement was false and misleading by omission, wouldn’t it?
Paul Patterson: I think the witness statement generally needed to be more comprehensive and it did – it absolutely missed those points that you’ve just alluded to and it would be misleading.
Mr Beer: Because it gives the impression of extraction essentially onto a CD?
Paul Patterson: Very simply, yes, it does.
Mr Beer: Can we please turn to the utility of the ARQ data and look at what might be one of the most significant paragraphs in your witness statement, which is paragraph 19. This is on page 11. You say:
“The Inquiry has asked Fujitsu to confirm whether, in its view, the ARQ data provided to [the Post Office] over time was sufficient to enable a postmaster to understand whether Horizon was operating correctly at their branch.”
Indeed, that is the question that we asked you to address.
“In the light of (i) the evidence heard by the Inquiry from postmasters during the Human Impact hearings, (ii) the evidence set out in the Fujitsu Witness Statements, and (iii) the matters set out in this corporate statement in relation to the ARQ Spreadsheet, Fujitsu cannot confirm that ARQ data on its own was sufficient to enable a postmaster to understand whether Horizon was operating correctly at the relevant branch in the time period covered by the ARQ data requested by [the Post Office].”
Paul Patterson: Correct.
Mr Beer: So you’re saying, by reference to three data sources there, or three bits of information or evidence, that the conclusion in the last three lines, “Fujitsu can’t confirm that ARQ data was sufficient to enable a postmaster to understand whether Horizon was operating at their branch correctly”?
Paul Patterson: Agreed.
Mr Beer: That’s, would you agree, a rather startling omission?
Paul Patterson: Yes, I would agree but, importantly, it’s the truth. On its own, the ARQ data could not give a subpostmaster all the data that they would need to determine whether everything in that environment was working correctly.
Mr Beer: Apart from the event logs that I’ve described, that weren’t accessed, the KELs that I’ve described, the notes left in the message store that were not accessed, has Fujitsu identified anything else that was needed to understand, from a postmaster and a court’s perspective, whether Horizon was operating correctly at their branch?
Paul Patterson: In our submission, in the company’s submission here, we haven’t identified any other material or any other systems that needed to be interrogated. In my reading of our documentation and given what I know, I think there are other areas that may well be – in terms of the other systems inside the Credence system and the POLSAP system, how these things all feed into each other – should also have been – should also have been checked.
Mr Beer: In the Group Litigation proceedings in the High Court in the Horizon Issues trial, the Post Office’s expert, Dr Worden, explained audit data as being central to the operation of the whole system, and he said it’s a central principle of Horizon that the core audit database acts as a secure gold standard for branches, a central principle of Horizon that the core audit database acts as a secure gold standard for branch accounts, and, indeed, the trial judge, Mr Justice Fraser, based, in part, a number of his findings on this evidence and what he was told about the centrality and completeness of ARQ data. I think you’ll know that from reading the judgment.
Paul Patterson: Yes.
Mr Beer: The issue that confronted the judge was why was ARQ data not sought in a number of the cases and the consequences for the reliability of action taken against the subpostmaster because ARQ data was not sought.
In the light of what you’ve said, I think you would agree that the provision of the ARQ data in the form that it was provided and the extent that it was provided was not really the gold standard at all.
Paul Patterson: No, it wasn’t.
Mr Beer: More a bronze standard or a copper standard?
Paul Patterson: I wouldn’t use that characterisation at all. I’ve seen –
Mr Beer: Pewter?
Paul Patterson: Well, I have seen one of the examples, I think, in Mr Castleton’s case and looking at that spreadsheet, and it’s, I think for me, it’s impossible to determine from there, and that’s certainly not a gold standard or any standard. It’s a very simple Excel file which tells you not very much.
Mr Beer: Do you know why Fujitsu allowed that mischaracterisation of the position to stand in the course of the Group Litigation?
Paul Patterson: No, I do not.
Mr Beer: Whether or not it was the gold standard, as it was presented in the Group Litigation, in fact, you, I think, agree that ARQ data was only a start –
Paul Patterson: I completely agree.
Mr Beer: – and, at the very least, ARQ data ought to have been provided in any case of the investigation or prosecution of a subpostmaster for a criminal offence?
Paul Patterson: Yes.
Mr Beer: Looking at the three reasons that you gave for ARQ data not being sufficient to enable a subpostmaster to understand whether Horizon was operating correctly at the relevant time, the first reason was evidence heard by the Inquiry from subpostmasters during the Human Impact hearings. What was it about that evidence that leads you to the conclusion that you’ve reached?
Paul Patterson: Listening to – or reading in my case, actually – the submissions, the notion that all of those subpostmasters had somehow all independently experienced the same thing and were all not aware – and couldn’t control it or didn’t know what was going on, is clearly not true.
There were problems and the subpostmasters were flagging those problems, and it is very clear from all of the evidence from subpostmasters that, on its own, it could not be relied upon: the ARQ data, that is.
Mr Beer: So it’s listening to the evidence of subpostmasters or reading the evidence of subpostmasters: how was that translated into “The ARQ data was not enough” conclusion?
Paul Patterson: So in the research and the work that the team have done in looking at all the evidence, all the commentary from the subpostmasters, to have that volume of commentary around the data, the appearance of it, how it was presented, how it was – there was clearly a problem in that process and that is why we’ve concluded it was not on its own sufficient enough for the subpostmasters to conclude.
Mr Beer: Thank you. The second reason you give for reaching the conclusion you do is the evidence set out in the Fujitsu witness statements. Now, the Fujitsu witness statements are the witness statements that were filed at the same time or roughly the same time as your corporate statement?
Paul Patterson: Yes.
Mr Beer: They included, for example, that of John Simpkins and Gerald Barnes, from whom we’ve heard this week. John Simpkins told the Inquiry in his witness statement and in his evidence, in relation to Legacy Horizon, that the message store provided a much more comprehensive account of the data held in the audit archive than did the ARQ data that was provided. Presumably it’s that kind of evidence that you’re referring to here as meaning that ARQ data was not good enough?
Paul Patterson: Yes, all three of the witness statements.
Mr Beer: Is that because, in this example, the information held on the message store would have been of use to subpostmasters who sought to challenge alleged shortfalls?
Paul Patterson: Yes, and I believe, in reading the witness statements from other Fujitsu colleagues, it was very clear that the message store was a far richer source of that information.
Mr Beer: The third thing you mention, as leading to the conclusion that Fujitsu has reached, is the matters set out in this corporate witness statement, yes?
Paul Patterson: Yes.
Mr Beer: Is that, essentially, the six subtopics that you go on to address in your witness statement from paragraph 26 onwards?
Paul Patterson: Yeah, it’s the – it talks about the table, yes.
Mr Beer: So if we look at the table –
Paul Patterson: Yes.
Mr Beer: – at the foot of page 14 –
Paul Patterson: Yes.
Mr Beer: – if that can come up on the screen, please, third corporate witness statement, page 14. It’s essentially those six things which are – can I call them problems with ARQ data?
Paul Patterson: Yes.
Mr Beer: That’s led you to the paragraph 19 conclusion too?
Paul Patterson: Yes.
Mr Beer: So let’s just see the context of this, if we go back a page to page 13. At the foot of the page, paragraph 26:
“The Inquiry has asked whether Fujitsu is aware of any cases where an ARQ log produced for the purposes of court proceedings against subpostmasters (i) may not have accurately matched the ‘original log files’, or (ii) was, or may be, unreliable (together [you call them] ‘ARQ reliability’).”
Then you tell us in paragraph 27:
“In the course of Fujitsu’s investigations to date, a number of incidents that may have impacted on either the underlying audit trail from which ARQ data is extracted or the ARQ extracts themselves have been identified. [Your] investigations have included both document searches and discussions with … employees. [They] are described in more detail below. Given the expansive period covered by the … requests, and the limitations of relying on and interpreting records of technical matters … without the benefit of guidance or explanation from relevant employees with contemporaneous knowledge, Fujitsu cannot be sure that these incidents contained in the witness statement are [exclusive]”, and if more are covered you will tell us.
Paul Patterson: Yes, correct.
Mr Beer: So just looking at the hierarchy of problems we are dealing with, one is that in the prosecution of subpostmasters, in many occasions, no ARQ data was asked or provided?
Paul Patterson: Yes.
Mr Beer: Top tier problem. Second problem, in the cases where ARQ data was provided, it wasn’t sufficient in itself, see paragraph 19 of your witness statement –
Paul Patterson: Yes.
Mr Beer: – and suffered from the defects that your employees have described?
Paul Patterson: Yes.
Mr Beer: Third tier problem, in any event, there were incidents across time that affected the very reliability of the ARQ data itself?
Paul Patterson: Yes.
Mr Beer: Can we look, then at that third tier problem: reliability of ARQ data, so, bearing in mind the health warning you give us here, that this shouldn’t be taken to be exhaustive or complete. But what you are telling us is, is this right, that the six incidents or six issues that you set out in the table do cast doubt on the reliability and completeness of ARQ data?
Paul Patterson: Yes.
Mr Beer: Can we look, please, at paragraph 28 and 29, on page 15. You say:
“… I understand that Fujitsu has identified approximately 2,400 ARQ requests dating from November 2002 onwards. For reasons highlighted … above, it has not been possible to conduct a forensic investigation into the ARQ reliability of the audit data supplied to [Post Office in respect of ARQ requests.”
Is that essentially because the raw data is not available?
Paul Patterson: In some cases, yes. I think before 2007, we don’t have the raw data.
Mr Beer: “The following summaries of incidents which Fujitsu has identified as having a potential impact on the issue of ARQ reliability have been prepared from documents produced to the Inquiry.”
They’re not within your personal knowledge.
So just translating the effect of those two paragraphs, 28 and 29, it’s right that the six issues that we’re going to speak about in a moment, referred to in the remainder of your witness statement, haven’t been run against the 2,400 ARQ requests that the Post Office made –
Paul Patterson: Correct.
Mr Beer: – to see whether they, in fact, afflicted that data. So secondly, is this right, the actual impact of the six incidents or issues on the 2,400 ARQ requests is not known?
Paul Patterson: In some instance, they are very time bound some of these six topics.
Mr Beer: Yes.
Paul Patterson: So they wouldn’t have applied to a previous version of Horizon. So there will be definitely – not all of them would have to be applied to all 2,400, because of time.
Mr Beer: Yes.
Paul Patterson: Other than that, you – I would agree with your characterisation.
Mr Beer: But looking at it the other way round, would you agree that six incidents or issues are not assessed to have had no material impact on the reliability or integrity or completeness of ARQ data otherwise you wouldn’t be telling us about it?
Paul Patterson: Yes, agreed, yes.
Mr Beer: Instead, they have at least the potential impact on the reliability of some of the ARQ data?
Paul Patterson: Yes.
Mr Beer: I think it follows from that, is this right, that you accept that the Post Office ought to have been told about the six incidents or issues?
Paul Patterson: Yes, they should have been and I believe they were.
Mr Beer: In relation to all six?
Paul Patterson: We might go through all of them now. As I said, I think, earlier, there was certainly a delay in some correspondence we’ve seen in evidence between when a problem which could affect was communicated, but I am unaware, Mr Beer, whether any of these were held back entirely.
Mr Beer: Okay, well, we’ll look at that as we go through each of them.
Paul Patterson: Okay.
Mr Beer: Just to be clear, so everyone can work out where we are, this is a completely separate issue from the 29 bugs, errors and defects. That’s errors and defects with the operation of the Horizon system. We’re dealing with six incidents or issues, problems, concerns, with the production of ARQ data?
Paul Patterson: Yes.
Mr Beer: But just going back to my question, I think it follows that you accept that the Post Office should have been told about these six incidents?
Paul Patterson: Yes.
Mr Beer: That’s because that would enable the Post Office to either decide whether reliance could be placed on the ARQ data that it was being given or not and, if it decided that reliance could be placed on the ARQ data, then it would need to tell the Defendant and the court about the issue or issues it had been told about by Fujitsu?
Paul Patterson: Agreed.
Mr Beer: So it can give proper disclosure of the flaw in the data, which might cast doubt on its reliability?
Paul Patterson: Agreed.
Mr Beer: That’s why they needed to be told. I think it’s clear from the evidence you’re going to tell us about a number of the six issues, that the incidents were indeed known about by the Post Office and discussions took place at a senior level between the Post Office and Fujitsu, including the consequences of the error and whether compensation needed to be paid by Fujitsu to the Post Office; is that right?
Paul Patterson: Compensation from Fujitsu? I didn’t quite catch the last part, sorry.
Mr Beer: Yes, when we look at one of the incidents, we’ll see that there was a threat of litigation from the Post Office to Fujitsu?
Paul Patterson: Was that the broken audit trail?
Mr Beer: Yes, it was.
Paul Patterson: Yes, indeed, I understand – so I agree with you, Mr Beer, on that one, yes.
Mr Beer: So some of them escalated up to a senior level?
Paul Patterson: Yes, I follow you now. Yes.
Mr Beer: Are you aware of any instances where there was a known or suspected issue with the ARQ data and that ARQ data was, nonetheless, relied on by the Post Office in a criminal or civil proceeding bought against a postmaster?
Paul Patterson: Have you got an example?
Mr Beer: I’m asking you whether, as part of the process that you’ve gone through, that you have, and your company has discovered, never mind the 2,400 requests that were put into us by the Post Office, let’s look at the ones that actually resulted in criminal proceedings and a conviction.
Paul Patterson: Mm-hm.
Mr Beer: Can we see whether any of these six bugs, I’m going to call them, afflicted the reliability of that data?
Paul Patterson: So my evidence says, yes, it could have done. Yes, it could have done.
Mr Beer: Can we look, then, at the six incidents, and the first of them comes on page 15 of your witness statement. I’m going to deal with these at relative speed because the detail is set out in your witness statement in detail, and is backed up by over 300 exhibits.
Firstly, the broken audit trail. This is problem number 1, I’m going to call it. We can see what the issue is on page 15 at paragraph 30:
“In or around May 2001, it was identified that there was a data loss in the audit trail for a six-day period in August 2000 …”
So I think that tells us that nine months after the data loss occurred, it was recognised, yes?
Paul Patterson: Yes.
Mr Beer: “At this time, audit data was gathered by an audit server and written to [a tape] for long-term storage, to be retrieved when needed. Two ‘data centres’ … at Bootle and Wigan, which contained the main Horizon servers.”
Then you go on to describe in more detail how the data loss occurred.
Then if we look at the second point, which is notification to the Post Office, that’s paragraph 33 on page 17. So paragraph 33:
“According to a letter from Jan Holmes … to Sue Kinghorn [so that’s from Fujitsu to Royal Mail] dated 23 May 2001, Fujitsu identified the Broken Audit Trail when undertaking an audit data extraction for [an Internal Crime Manager in the Post Office] in relation to an ARQ request [number 8]. It appears that data for the period 8 to 14 August … was held on four [of the] tapes.”
Then if we go over the page, please, to page 18, paragraph 35. You’re quoting, in part, from the letter.
“On 9 May, [Post Office] was notified that Fujitsu was unable to source the evidential data requested.”
Then in the letter you’ve just referred to:
“[Fujitsu] informed [the Post Office] of the issue and explained that ‘the break has arisen due to a combination of events out side [Fujitsu’s] immediate control but it does mean we are not able to retrieve TMS records for that 6 day period. All other elements of the audit trail are complete’.”
Then if we go to page 24, please. In paragraphs 40 to 51, you have described for us the attempts to recover and rebuild that missing data but the essence of it is in this paragraph 49, here:
“By October 2001, the … back-up tapes had also been recovered from the relevant data centre … In order to reconstitute the audit data from the … back-up tapes, a pseudo audit server was built which the back-up tapes were loaded on to … by … December 2001, Fujitsu had reportedly identified that about 66% of the missing data … was available on the back-up tapes. The remaining 34% was not present … and was deemed irretrievable. The gap in the audit trail was therefore said to have been reduced from a period of 6 days … to less than 24 hours …”
Paul Patterson: Correct.
Mr Beer: So is that an example of what you would describe as a limitation in time of the affect that this issue had on the audit trail?
Paul Patterson: Yes.
Mr Beer: Then, if we go forward to page 27, please. You address steps taken to prevent this happening again. I’m not going to go through all of those but, if we go forward to page 31, please, at paragraph 63 – in the preceding paragraphs, you have dealt with essentially an escalation of correspondence between Fujitsu and the Post Office, concerning allegations of breach of contract by Fujitsu and threats of litigation by the Post Office, and the outcome of this was that Post Office and Fujitsu agreed to settle any claims regarding possible breaches by Fujitsu of its contractual obligations in return for a payment of £150,000.
Paul Patterson: Yes.
Mr Beer: Can we turn to problem number 2, please, “Omissions in ARQ Data Caused by Operator Error”. You deal with that immediately underneath these paragraphs here and describe the issue in paragraph 64.
So we’re moving forwards in time here to 2003. Data had been omitted, it was discovered in 2003, in response to three requests – and you give the numbers – related to Forest Gate, and one request related to Urmston, and that is in July and December 2003, respectively, yes?
Paul Patterson: Yes.
Mr Beer: Essentially, you go on to describe a series of operator errors that occurred when the operator was seeking to recover data from those two branches in respect of those four ARQs?
Paul Patterson: Yes.
Mr Beer: If we go forwards to pages 34 and 35, in paragraph 72 you explain the explanation given to the Post Office at the time of the cause of the omissions, yes?
Paul Patterson: Yes.
Mr Beer: I should have said a moment ago that these omissions were picked up by a change in personnel, is that right, who was going to attend court? So the person that had originally conducted those four ARQs couldn’t attend court. A new Fujitsu employee was bought in to attend court, Penny Thomas, I think, and she reran –
Paul Patterson: Yes, that’s correct.
Mr Beer: – the four ARQs and found that there was missing data?
Paul Patterson: Correct.
Mr Beer: So, essentially, that was by chance?
Paul Patterson: Yes.
Mr Beer: If the original Fujitsu employee had been able to attend court, this wouldn’t have been discovered.
Then we go forward to the conclusion on page 35, paragraph 74. Mr Mitchell provided a witness statement in relation to the two branches and he concluded that the omissions made in the data provided by Ms Lowther (that’s Neneh Lowther, the first person, the one who turn up at court) had not been provided in the data provided by Penny Thomas and that the latter data was complete in accordance with the original ARQ.
Paul Patterson: Yes.
Mr Beer: So that’s a case where it was revealed and revealed in a witness statement?
Paul Patterson: Yes.
Mr Beer: The third problem, please. The Riposte lock event of 2008. You describe this at the foot of the page, page 75. We’re moving forwards now to December 2007. An incident was reported by a branch to the Network Business Support Centre, that’s a POL operation. It was recorded in a PEAK and referred to Fujitsu. The information was “That a BM stock unit had a gain of £465.73” which “did not go to local suspense”. When the stock unit rolled over, the local suspense was cleared and the gain was not accounted for. The value of the gain was shown on the trading position line on the branches trading statement. The trading position line “should always show zero”.
Now, in the following paragraphs, 76 to 109, you set out the history of this Riposte lock event, which that that effect there, and how it was first identified by Fujitsu in 2007, and then extensively considered during 2008, during a series of calls, emails and meetings culminating in an internal email and presentation to Fujitsu employees on 17 December 2008. So if we go forward to paragraph 110, please, which is at the foot of page 49, this is about when the Post Office were told.
“On 7 January 2009 a Fujitsu employee notified Sue Lowther of Post Office and David Gray of Post Office about the 2008 ARQ issue via email. Ms Warham provided a summary of the 2008 ARQ issue and similar terms to that set out in a proposed witness statement, which we’ve previously narrated, and the “various steps that should be taken by Fujitsu and Post Office to address the issue, including”, and then you set them out between (a) and (e).
Paul Patterson: Yes.
Mr Beer: Given it was clear to Fujitsu in 2008, throughout the course of 2008, remembering the incident had first been notified in December 2007, that the Riposte lock was an issue capable of impacting on criminal and civil litigation, for which ARQ data was being requested and provided, do the papers that you’ve read or the briefings you’ve received reveal why Fujitsu didn’t alert the Post Office to the issue immediately –
Paul Patterson: No.
Mr Beer: – and not seemingly, for the first time, until 7 January 2009?
Paul Patterson: So the evidence we’ve got is that it was delayed and I don’t know why.
Mr Beer: The papers don’t reveal why?
Paul Patterson: No.
Mr Beer: Would you agree that, as the issue was one which was capable of impacting on criminal and civil proceedings, as is later recognised, it ought to have been notified to the Post Office promptly?
Paul Patterson: Yes.
Mr Beer: Can we just look, please, at the email that you refer to here in paragraph 110. If we just go back, please. At the foot of the page, you refer to the email on 7 January, Ms Warham notified Sue Lowther via an email, and it’s your footnote 175.
Can we look, please, at FUJ00155399. It’s the Wendy Warham email at 10.46, the bottom two-thirds of the page, thank you:
“… I have left you a voicemail as I need to update you on a recent issue that has occurred and has been resolved but does have some short-term impacts.”
Would you agree that’s being presented by Fujitsu as an issue that’s already been solved and fixed. It’s only recently arisen, “a recent issue”?
Paul Patterson: Well, that’s exactly what it says.
Mr Beer: Whereas, in fact, this dated from December 2007, didn’t it?
Paul Patterson: In the earlier – where we saw the problem, yes, it did.
Mr Beer: Going back to your paragraph 111, please, which is on page 50 of your witness statement, you say:
“… Ms Thomas forwarded Ms Warham’s [witness statement] to Dave Posnett [in Post Office] attaching the proposed witness statement.”
Can we just look at that witness statement please, it’s FUJ00122604. This is the draft witness statement that it’s proposed by Fujitsu is going to explain, in any court proceedings, the Riposte lock event. Can we just look at page 7, please. At the top of the page, so if we can look at the top half:
“In December 2007 an occurrence was reported in one office … This led to a previously unseen database lock where an administrative balancing transaction failed to be written to the local message store database. This generated a generic and non-specific software error … A financial imbalance was evident and was subject to investigation by [Fujitsu] and Post Office. The financial imbalance has been resolved.
“A software correction was applied across the estate in early November ‘08 to ensure that any such event generated would be monitored. Testing of the correction has established that the unmonitored error does not occur elsewhere in the system.”
It’s proposed by this witness statement, would you agree, to summarise the issue by offering reassurance that testing confirmed that the issue did not occur elsewhere in the system?
Paul Patterson: Yes, that’s what it’s trying to do. Yes.
Mr Beer: So, effectively, reassuring the Post Office that this was an isolated incident, not affecting any other case in which ARQ data had been supplied?
Paul Patterson: That would be the inference from this, yes.
Mr Beer: That’s not accurate, is it, on the information that you have been provided with?
Paul Patterson: Because it lasted longer?
Mr Beer: Yes, well, the fix in November ‘08 is about monitoring the event after that time. What’s overlooked is the operation of the event from December ‘07 until November ‘08.
Paul Patterson: Before, yes, I agree. But in paragraph 1.10 I thought, Mr Beer, Ms Warham laid out going back and checking some other factors before, which I thought was the – point A.
Mr Beer: Yes, checking ARQs to confirm data integrity in the period May ‘07 to November ‘08.
Paul Patterson: Yes, yeah. Now, what I don’t know is whether that took place, which I think is what you’re –
Mr Beer: And, if it did, whether it ever got reflected in the witness statement?
Paul Patterson: Yes.
Mr Beer: Because we know what happened is it was decided, in the event, not to reveal any of this in a witness statement?
Paul Patterson: Correct, the advice – well, you might be going to this – was to remove those two paragraphs.
Mr Beer: The Post Office lawyer’s view, Rob Wilson’s view, was that wasn’t necessary to give disclosure of this incident but he said, “if we’re sure that there have been no other incidents”. Do you know what steps were taken to determine whether there had been any other “incidents”, as he called them?
Paul Patterson: No, I don’t.
Mr Beer: Are you aware overall of the provision by Fujitsu of information of the outcome of any investigation as to whether the 2008 data lock had affected any other ARQs?
Paul Patterson: So I thought there was – it may not have been this particular problem but I think there is something later on in our answers around having checked a number of ARQs where problems had been presented or not. I think in the case of this 2008 one, I do not know whether previous ARQ material had been checked against that problem.
Mr Beer: What about, irrespective of ARQ data, whether data had been checked in that period to see whether the Riposte lock had afflicted the integrity of that data?
Paul Patterson: I’m unaware of any other checks that had taken place.
Mr Beer: We do have an email, if we can look at that please, FUJ00155421. Thank you. This is February, it’s the second email on the page, 2009, from Penny Thomas.
Just scroll up, please, to Dave Posnett in Post Office:
“… analysis of the data covering [May ‘07 to end November ‘08] has been completed.”
Second line, next paragraph: 27 instances of concern were identified, they have been fully analysed. We can confirm that the locking was caused by contention between the end of day process and a Riposte checkpoint being written. No transactions or balancing activities were affected.
So I think that answers the first point that we were looking at, whether there had been a back check over that relevant period.
Paul Patterson: Yes.
Mr Beer: But that’s in relation to 195 ARQs; it’s not in relation to data that hadn’t been the subject of an ARQ?
Paul Patterson: Correct, and that’s what I was referring to, Mr Beer. I said I thought that we had – there had been some analysis back then, going back to the ARQs that had been submitted.
Mr Beer: Could we turn to the duplicate transactions issue, which is problem number 4. This is paragraph 118 of your witness statement, which is on page 54. You tell us in 117:
“The Inquiry [as indeed we did] has asked for details of the Duplicate Transactions incident, first identified in 2010 [and] recurrences of [it] in 2014 and 2016.”
You describe what the incident was, in paragraph 118 onwards. I’m not going to rehearse that; it’s quite straightforward.
If we go forwards to 131, please, which is on page 61. You indicate in paragraph 131 that Post Office were informed about the incident on 30 June 2010.
Paul Patterson: Yes.
Mr Beer: I’m going to summarise the relevant email. It stated that Fujitsu had identified the affected ARQs. It had already developed a workaround and the workaround would enable anyone looking at ARQ data to identify the duplicate transactions and ensure that they weren’t brought into account.
Paul Patterson: Yes.
Mr Beer: Presumably, the idea of that is to reassure the Post Office that the data still had integrity, if you used the workaround, you could still rely on the data?
Paul Patterson: Yes.
Mr Beer: If we go forwards, please, to paragraph 137 of your witness statement, which is on page 64. Between 137 and 140, you refer to the fact that the Post Office requested a witness statement to explain the issue and the workaround.
Paul Patterson: Yes.
Mr Beer: I’m summarising.
Paul Patterson: Yes, sorry, yes.
Mr Beer: And that, subsequently, Fujitsu provided such a witness statement?
Paul Patterson: Yes.
Mr Beer: Have your investigations and your briefings in the documents you’ve read revealed to you whether the statements provided by the Fujitsu witnesses in relation to data afflicted by the duplication issue are true and accurate?
Paul Patterson: So I think we’ve got an example in here. I find, personally, the language very convoluted, rather than being very, very clear, but there was – there is a statement in the witness evidence trying to describe what had happened with the audit.
Mr Beer: But your reading of the material, is this right, is that, as far as Fujitsu was concerned, the ARQ data could continue to be relied upon, so long as the workaround had been deployed?
Paul Patterson: Yes.
Mr Beer: Therefore, it was reasonable for the Post Office to continue to rely on those data?
Paul Patterson: That would have been my understanding of it, yes.
Mr Beer: Can I turn to issue 5, please, historic gaps in the ARQ data. You deal with this in paragraph 153. I think this is new material – sorry, it’s on page 73. This is new material for the Inquiry, so it wasn’t something that we were asking Fujitsu about that we already knew from reading the primary material. This is a voluntary revelation:
“… Fujitsu has identified a small number of documents in relation to this topic …”
You describe what happened from paragraph 154 onwards. Is it right, in relation to this incident, that there had not been revelation to the Post Office or was it essentially addressed in the course of meetings?
Paul Patterson: The way I understand what the emails were saying, which is towards the end of that paragraph, Mr Gauntlett was asked to write up and communicate to (unclear) in meetings, so exactly to the point you’ve just made.
Mr Beer: Then, lastly, please, if we can look at incident number 6., bottom of page 78:
“In addition to the five other issues … I understand that Fujitsu disclosed to the Inquiry 102 documents from its PEAK, PinICL and Known Error Log databases … These have been identified … as records of incidents referring to the ARQ process in the context of court proceedings.”
Over the page:
“Many of these documents relate to system changes and support issues … However, there are certain PEAKs set out below that could be relevant to the issue of ARQ reliability.”
But in the course of time available those eight – I think there are in total – PEAKs haven’t been analysed; is that right?
Paul Patterson: Correct.
Mr Beer: Does that remain the case?
Paul Patterson: Yes. As far as I know, Mr Beer. As far as I know, yes it does.
Mr Beer: Thank you. That can come down.
Would you agree that that collection of six incidents presents a rather sorry tale of the reliability of ARQ data?
Paul Patterson: Yes, it does. There were many, as is the evidence we’ve just been through, many bugs and errors throughout of the time with the Horizon system in all three of its – or certainly in its first two incarnations. So, yes, I would agree with you.
Mr Beer: That is within of the three building blocks you’ve used to reach the conclusion that you have in paragraph 19 –
Paul Patterson: Yes, it is.
Mr Beer: – that ARQ data ought not to have been relied upon and presented as being an accurate and complete record of the health of the transactions conducted by a subpostmaster at his or her branch?
Paul Patterson: Agreed, and I would expand it slightly to say, I think, just for the overall health of the system, which I think was one of the earlier questions, I think for the subpostmaster that data on its own is not sufficient.
Mr Beer: Are you aware of an experienced Fujitsu employee from the SSC being required to give evidence in court in 2006, Anne Chambers, and writing after the event a document that sought to ensure that problems that she had encountered –
I’ll just stop there. Sir, you’ve disappeared from the screen. Can I check that that does not mean that you are not connected?
Sir Wyn Williams: I’m still here and I am able to view everything that’s going on in the hearing, so it’s –
Mr Beer: It’s only us that’s losing out?
Sir Wyn Williams: Yes.
Mr Beer: I’ll continue with the –
Sir Wyn Williams: Good way of putting it, Mr Beer.
Mr Beer: I’ll continue with the question.
I was asking you about Anne Chambers giving evidence in civil proceedings, the Lee Castleton civil proceedings, in 2006.
Were you aware of her preparing a document about problems that had arisen in the course of preparation for and the act of giving evidence in those civil proceedings?
Paul Patterson: So in the pack there is – it’s one of the – in the supplementary pack there is her document that you’re referring to, which I have –
Mr Beer: “Afterthoughts” document?
Paul Patterson: Forgive me?
Mr Beer: It’s called “Afterthoughts”.
Paul Patterson: “Afterthoughts”, yes. So I have read that document.
Mr Beer: Can we look at that please, FUJ00152299. Can we see from the second page at the foot, it’s signed off by Mrs Chambers on 29 January 2007. If we go back to page 1, she deals with four topics: the initial approach to SSC staff, the review of the technical evidence that had been undertaken before she was called to give evidence; the disclosure of evidence into the proceedings; and accessing Helpdesk calls.
Before we gave this to you recently, was this a document you were aware of?
Paul Patterson: No.
Mr Beer: If I just refamiliarise you with it before I ask some questions. She says in summer 2006 she was asked by the Security Manager whether she would be prepared to speak to a solicitor about a call she’d dealt with in 2004. Her initial response was this wasn’t normal process:
“… he reassured me that it was more or less a formality so somewhat reluctantly [she] agreed.
“Subsequently, before meeting with the solicitor, he asked me what [her] availability was in the autumn for the court case. This was the first mention there was of any possibility of going to court. Repeated reassurances that this would all be settled before getting court proved to be unfounded.
“… there may be circumstances where witnesses are summoned and have no option but to comply, but I was not at all happy about to be how this was handled.”
At this time, are you aware of, within Fujitsu, how approaches to witnesses to provide either evidence of fact or expert evidence were supposed to be handled?
Paul Patterson: So I only know what I’ve seen in my document because I wasn’t there at the time.
Mr Beer: Yes.
Paul Patterson: As someone in my position now, I’m surprised that that was even taking place, where there was a direct connection between a member of staff and a solicitor representing a subpostmaster.
Mr Beer: No, it’s a solicitor representing the Post Office.
Paul Patterson: Sorry, yes, representing the Post Office. I am surprised that that was even part of the day-to-day working.
Mr Beer: Her second issue, “Review of technical evidence”:
“When I took the initial call in February 2004 …”
What she’s referring to here is when she was doing her day-to-day work in the third line support, she took a call and she administered it on the relevant PEAK:
“I spent only a few hours on it before deciding that I could not see any sign of a system problem. I looked only at a couple of weeks’ information.
“While in this case I am now sure that I did not miss anything, and my initial analysis was correct, I am concerned that there was no technical review of the Horizon evidence between the original call and the case going to court. It is probable that any system problem affecting the accounts would have shown up to Post Office staff who did check the figures very carefully, but since the postmaster was blaming the system for the losses I think it would have been sensible to have double-checked this within Fujitsu before it got as far as court. I was certainly concerned, in the early stages, that there might be something I had missed.”
Just stopping there, she’s making a suggestion that in investigation or prosecution or in civil cases, one doesn’t just, before taking action against the subpostmaster, rely on what was done back in the day; there’s a recheck done. To your knowledge, was what she suggested there ever implemented?
Paul Patterson: From my knowledge, no, it wasn’t, and I think her suggestions are very, very important. These are serious matters which isn’t just ticking a box, and I think the point she’s making here is that, given the seriousness of it, further analysis should have been done to make sure it was correct and all the data was correct.
Mr Beer: “Once in court [she continues], I found myself being treated as an expert witness and answering a wide variety of questions about the system, although nominally I was a witness of fact and my witness statement covered just the investigation done in 2004. Fortunately I do have extensive knowledge of the system and was able to fulfil the wider role – but what would have happened if the initial call had been handled by a less experienced SSC person?
“If there is a similar case in future, where the system is being blamed, would it not be sensible to have a technical review of all of the evidence, at the first indication that a case may be going to court? Someone involved in that review would then be well placed to give evidence in court.”
I think you’ll agree, again, that’s a very sensible recommendation.
Paul Patterson: A very sensible recommendation and I’ve learnt, in preparation for the Inquiry, the change – the term “expert” and “witness of fact”, which is very, very important.
Mr Beer: We saw in the 2005 process document that a distinction was drawn between a person providing a witness statement of fact and a person being called to provide expert evidence but that document did not precisely delineate the difference between them –
Paul Patterson: No, it did not.
Mr Beer: – and it didn’t explain to Fujitsu staff the boundaries of each and the additional duties that apply if one is giving expert evidence?
Paul Patterson: Agreed, Mr Beer, agreed.
Mr Beer: Her third point, “Disclosure of evidence”:
“Fujitsu made a major legal blunder by not disclosing all of the relevant evidence that was in existence. I found myself in the invidious position of being aware that some information (Tivoli event logs) existed, but not sure whether they had been disclosed or not, since I had not been party to any of the requests for disclosure. It came evident in court that they had not been disclosed”.
She quotes from an email from the Post Office’s solicitor after the revelation and then, scrolling down:
“This suggested that the disclosure of the message store itself was an afterthought, though it is fundamental to the system.”
I think that reflects something you have told us already?
Paul Patterson: Yes.
Mr Beer: “I know that for fraud cases the ‘transaction log’ and ‘event log’ are extracted from the full message store and submitted, but surely the full message store has to be disclosed in all cases?
“Many other files are also archived to the audit server as a matter of course and could hold relevant information, although the Security Team are not necessarily aware of their existence or potential relevance. I’d like to suggest that a list of these files is compiled so that similar mistakes are not made in the future.”
Again, sensible recommendation, you’d agree?
Paul Patterson: I would agree and, when I read this, I was – sadly, again, a word I’ve used before – appalled that we have a solicitor writing to a member of staff pointing out obligations and I think, as we’ve already discussed in evidence, the material needed – rightly needed by the subpostmaster needed to be far more comprehensive, which is what Mrs Chambers is alluding to here.
Mr Beer: It’s essentially a 2007 reflection of some of the points that you make in paragraph 19 of your witness statement 17 years later?
Paul Patterson: Yes, it is, Mr Beer.
Mr Beer: “And what about calls on PEAK, which may have evidence attached? And any evidence which may have been kept within the SSC? I was not asked whether I had anything that might have been relevant …”
As it happens, she didn’t:
“Of course there may be subtleties to this that I am unaware of, whereby data may exist but there is no obligation to disclose it. If this is the case, could any future witnesses be briefed appropriately? The response ‘no one has ever asked for that before’ does not seem to be a good reason for non-disclosure.”
I think you would agree with those sentiments?
Paul Patterson: I would agree with those sentiments, yes.
Mr Beer: Lastly, “Helpdesk calls”, her fourth issue:
“The case highlighted a common problem, both in 2004 and now. The postmaster raised many calls about his continuing losses, both with Horizon and with the NBSC. These kept being bounced and it took weeks before a call was passed to SSC.”
We’ve heard a lot of evidence about this, a merry-go-round as between the NBSC and the Helpdesk:
“Strictly speaking, problems with discrepancies do need to be investigated by NBSC in the first instance, but where there are continuing unresolved problems it should be possible to get the issue investigated properly, and one of the Helpdesks should be prepared to take responsibility … Personally I think … the Horizon Helpdesk is penalised for passing ‘Advice and Guidance’ type calls on to third line [because this] leads to do many calls being closed without proper investigation or resolution. This is very frustrating for postmasters, though possibly not an issue of concern to [the Post Office].”
Again, sensible and reasonable advice about an earlier escalation within the tiers of the Helpdesk system?
Paul Patterson: Yes, I think Ms Chambers has made many – several points in this document which are very sensible, and I think for a subpostmaster who is constantly calling up for problems, for that not to be flagged as to needs an intervention and a proper intervention, not just passed around, it was wrong.
Mr Beer: So would you agree, overall, that this document lists a series of opportunities, each of which was missed by Fujitsu?
Paul Patterson: I don’t know what action was taken from this but, as I said a moment ago to you, I am unaware of whether this is – did it go into day-to-day operations or not?
Mr Beer: We’ve heard evidence that the response to this from Mrs Chambers’ manager was essentially a pat on the head and said, “Thank you very much”, and then they carried on with business as usual.
Paul Patterson: Well, that’s wrong, Mr Beer. That’s wrong.
Mr Beer: If that evidence is accurate and nothing was done in relation to each of these four issues, would you agree that this is a series of missed opportunities?
Paul Patterson: I would agree it is a series of missed opportunities.
Mr Beer: Thank you very much.
Mr Patterson, they’re the only questions that I wish to ask you today. As I’ve said already, there will be more extensive questions of you on a broader range of issues when you come back to give evidence later in the year. But thank you very much.
I don’t know whether there are any questions from representatives. I think there are, sir, so might that be an appropriate moment to break until 2.00 pm?
Sir Wyn Williams: Yes, certainly.
Mr Beer: Thank you, sir.
(1.00 pm)
(The Short Adjournment)
(2.04 pm)
Mr Beer: So first, it’s Mr Stein.
Questioned by Mr Stein
Mr Stein: Mr Patterson, my name is Sam Stein, I appear on behalf of many subpostmasters and mistresses and I’m instructed by a firm of solicitors called Howe+Co.
Now, online we can see that Fujitsu publishes its purpose, which is to “build trust in society through innovation”. Do you agree that, instead, the combination of Fujitsu and the Post Office’s creation of this scandal has damaged trust in innovation within society?
Paul Patterson: I think the history of this appalling miscarriage, Mr Stein, would tend to point to, yes, trust has been broken in technology and how technology is used and not used. It has tested us all.
Mr Stein: Mr Patterson, you’ve explained this week and today that Fujitsu is committed to engaging with government to provide a suitable contribution and redress to subpostmasters and their families. Within this Inquiry, the question of how compensation is being dealt with has been described variously. We’ve put it as being a “scandal within a scandal”.
Will you commit to making sure that what Fujitsu does regarding its own offer to put money into the Government’s pot to assist with compensation does not delay or sidetrack the efforts to secure proper compensation for subpostmasters and mistresses.
Paul Patterson: Mr Stein, yes, I can commit to that and I think the way that we’ve tried to engage with the Inquiry and engage with Government shows that we are very committed to finding the truth, and we will do nothing – nothing – to delay that.
Mr Stein: Now, you’re aware, Mr Patterson, that subpostmasters, mistresses, managers, people who worked in the branches, their lives and families’ lives have been ruined, devastated by the decades of this scandal. Those people that we represent have lived through the Great Post Office Cover-Up, the Second Sight investigation, Second Sight and what happened in relation to their reports, the time that the brave 555 subpostmasters took on the Post Office at the High Court and exposed the corruption within the heart of this scandal, and then the criminal Court of Appeals, and now, of course, they’re living through the Post Office Inquiry.
Help us understand, Mr Patterson, why did it take so long for Fujitsu to decide it had a moral obligation to put its hands in its pockets and provide money, not just apologies?
Paul Patterson: As you said in your comments, this is a decades-old miscarriage, which started a long, long time ago and involves many, many people and organisations in that. I think Fujitsu, more recently, as we have understood more, we have clearly let society down and the subpostmasters down. I think we had our obligations to the Post Office to be at the front of everything we were doing and that was wrong, and I think, subsequently, we’ve now seen where the evidence has taken us and the investigation has taken us, and that’s why you have had the statements from Fujitsu more recently.
I can’t comment on the past, Mr Stein. I don’t know why things weren’t done in 1999 or 2005. I don’t know. But what I can say is that, as long as I’m here and doing what I’m doing, I’m going to do everything to make sure we get to the truth.
Mr Stein: But, Mr Patterson, that’s not quite true, is it? Fujitsu’s commitment, what it now appears to accept via you, a moral obligation to pay up, that’s only been most recently expressed by you, whereas the apology from Fujitsu has been expressed in the opening stages of this Inquiry, in your statements to this Inquiry. Why has it taken so long for Fujitsu to own up to the fact it needs to put its hand in its billion-pound pockets and pay something towards compensation; why so long?
Paul Patterson: Mr Stein, I don’t know the answer to that question. I can only tell you my honest view today, which is we have come to this conclusion, as we’ve gone through the Inquiry and gathered our evidence and considered our moral obligations to the victims of this crime, and that is why you have heard what I’ve said most recently. Again, I can’t – I don’t know why in the past we didn’t say those things.
Mr Stein: Have you met a single subpostmaster?
Paul Patterson: No, I have not, Mr Stein.
Mr Stein: Will you commit today to meeting with subpostmasters, their legal representatives, Howe+Co, the firm that has instructed me in relation to this matter, and discussing with them the way forward that Fujitsu can assist in seeking to show that Fujitsu means what it says by its apology and by its commitment to provide financial redress?
Paul Patterson: Mr Stein, I have not met any subpostmasters in the past because I didn’t feel it was appropriate for me to do that. If that is a request from the subpostmasters and their representation, I will absolutely do that and I will sit down there with my colleagues, as you’ve just said.
Mr Stein: You’re aware that the effect of this scandal upon subpostmasters was very much felt by their families, by their loved ones, by their partners, by their children. This devastated not only the subpostmasters’ individual lives but everyone around them. Subpostmasters, because they could not see what was going on, would sometimes blame their own partners or their employees, and the effect, therefore, of this scandal is much more widespread than on the subpostmasters themselves, whose lives sometimes have been terminated by this scandal.
Will Fujitsu commit to providing funds, in recompense for all of those hurts to others outside of subpostmasters and the compensation scheme, in other words financial redress to support others that have been affected? You may want to think that what could be done by Fujitsu is supporting people in the future, subpostmasters in future, entrepreneurial endeavours their families in such or in education. Will Fujitsu consider that type of support?
Paul Patterson: Mr Stein, the Inquiry and the exploration of the history of all of this is getting a great deal of attention and support from us. Our company here in the UK has many things in society around the country, around the UK. I think the suggestion that you’ve just made, and clearly in a conversation, if I was to be able to engage in that with subpostmasters and their representation – their representatives, would be absolutely something we would like to consider.
I think skills in our country, without jumping too far, is very important and I think there are things that we can do in our technology world that may or may not be of help to subpostmasters and their associated families.
So I would engage in that conversation, Mr Stein. I can’t promise the outcome of that but I would engage in it in the right way.
Mr Stein: Do you agree that a restorative justice process for subpostmasters, their families, their loved ones and everyone affected should include making sure that Fujitsu does what it can to put those people back into their rightful place within their own communities, restoring the trust and the honour that they have lost through this process?
Paul Patterson: Mr Stein, we are completely – the company is completely committed to doing that through the Inquiry and my comments to the Select Committee and to Government. There are many parties involved in this and we have a part to play in that, and we will do everything we can do to address those wrongs in whatever way we can.
Mr Stein: Mr Patterson, sitting to my left is Mr Enright, one of the partners at Howe+Co solicitors. He is available today for you to speak to. Can you make time to speak with him, to start that conversation with him today?
Paul Patterson: So, Mr Stein, we’ll finish at whatever time we finish today and I will be welcome to exchange contact details at least with Mr – from Howe+Co.
Mr Stein: Mr Enright?
Paul Patterson: Forgive me, Mr Enright. More – time – I don’t know when we’re going to finish but would I be willing? Absolutely, Mr Enright, and we can exchange our cards and we can have that conversation. Clearly, there are other people on Fujitsu’s side who would need to be attending to that.
Mr Stein: You’re probably aware, Mr Patterson, that some of our clients are here today. They may wish to speak to you or exchange details or set up meetings with you and, again, would you accept that that is something you will conduct, if possible, today but otherwise make arrangements with you and your team?
Paul Patterson: Mr Stein, I will talk to anybody at any juncture. I have a great deal of respect for the subpostmasters. I have no understanding because I’ve never gone through what they’ve personally gone through, and through our own team here and our representatives, more than happy to engage in that conversation. I have to say, I’m not sure today is the right time and the place for it, if you’ll allow me, but I’ll absolutely do that.
Mr Stein: Thank you, Mr Patterson.
Sir Wyn Williams: Who is next?
Mr Beer: Sir, I think Ms Dobbin has some questions.
Sir Wyn Williams: Yes.
Questioned by Ms Dobbin
Ms Dobbin: Mr Patterson, I hope you can hear me. I represent Gareth Jenkins, my name is Clair Dobbin.
I have a couple of questions for you, they’re rather more technical in nature. Are you in a position to explain or provide any information as to whether or not the Post Office, in fact, agreed the parameters of the ARQ return? Do you understand my question?
Paul Patterson: I think I do, Ms Dobbin. In my evidence today and I think previously, there’s a very clear document that says what the material should be – what the parameters, forgive me, inside an ARQ need to be and that was something that came from the Post Office to Fujitsu.
Ms Dobbin: Thank you. Do you happen to know why Post Office wanted that type of information or applied that criteria to what they wanted when they made an ARQ request?
Paul Patterson: Not all the time. I think in the flowchart we saw earlier, some of it would have been for prosecutory type support. I think, on the other side of that flowchart, was other reasons for ARQs and there could be a vast array of those, I’d imagine, that the Post Office criteria – why the Post Office would request that.
Ms Dobbin: All right. So it sounds, Mr Patterson, as though that question might be directed better to someone else who might understand those parameters. But thank you.
When we look at some historic ARQ requests going back quite a long time, and I’m talking about the pro forma type form that you will have seen that Post Office sent, have you seen any of those?
Paul Patterson: Yes, yes, they’re in the exhibits, I believe.
Ms Dobbin: It’s right, isn’t it, that one of the things that Post Office could request in that was confirmation that there were no reported system malfunctions; is that correct? Have you seen that?
Paul Patterson: I think it is in the exhibit, if we could probably go to the exhibit. I think it shows a criteria.
Ms Dobbin: I’m not sure have the exhibit to hand. I’ve certainly got one that maybe I could use, unless Mr Beer does have the exhibit.
Mr Beer: I think if we display FUJ00152562. That is an example one. I don’t know whether that will suit Ms Dobbin’s purpose.
Ms Dobbin: I’m very grateful to Mr Beer. So we have, I think this is a typical one, and – no, this is a different type of one. If I perhaps can try another type, I think these may have changed over time, Mr Patterson. Can I just see if I can get one that I think has been seen before.
May I try and see if POL00054125 is on the system. It’s not.
Mr Beer: That’ll be a no, it’s not on the system.
Ms Dobbin: Can you take it from me, Mr Patterson, it’s the ARQ request that was made in the case of Mrs Seema Misra and we see this in other ARQ requests made before that time in the standard form. Just to repeat, one of the options that allowed the Post Office was to ask for confirmation that there were no recorded system malfunctions. I think you’re not familiar with that, I think –
Paul Patterson: No.
Ms Dobbin: – from what you’re saying.
Paul Patterson: Correct.
Ms Dobbin: If that’s right, that obviously indicates, doesn’t it, that there was that option open to Post Office when making or giving – or when submitting an ARQ request?
Paul Patterson: Yes, if that was one of the options on the submission, yes.
Ms Dobbin: Thank you.
I wanted to go, then, to your witness statement and, again, Mr Patterson, if I’m asking you a question that’s beyond your knowledge, please do say. So if we go to your witness statement at WITN06650300, and please may we go to paragraph 102, page 47. Thank you.
Now, this is talking about changes that were brought in in the context of Craigpark, and that’s talking about event messages.
Paul Patterson: Okay, yes.
Ms Dobbin: Have you had a chance to look at that?
Paul Patterson: Yes.
Ms Dobbin: Then paragraph 103, please. This is talking about a change proposal that was made in relation to the audit system, and this is talking about the automatic retrieval and recording of event data. As I’ve said, please, if this is beyond your knowledge, please do say, but is it right that the checking of NT events that that started to become part of a routine process after the Craigpark incident?
Paul Patterson: So, you’re correct: I don’t know.
Ms Dobbin: You don’t know. All right. Will try someone else with that. Thank you, Mr Patterson.
Then, finally, just this: when you were asked questions about Anne Chambers and her experience at the High Court, you said that you were horrified that she was being contacted directly by a solicitor who was acting for the Post Office and you appeared to be basing that on your knowledge of how a corporation should work and how it’s employees should be involved in legal proceedings; have I got that right?
Paul Patterson: So my concern was that any employee was involved in a direct conversation with a legal representative without representation from either our company or our own legal representatives. It’s such a serious engagement, I felt it was not appropriate and, clearly, Mrs Chambers didn’t feel the same, for her just to be involved in that conversation.
Ms Dobbin: So, in other words, the company would act as a sort of interlocutor to be part of the dialogue with the employee to make sure they understood what they were doing and what was being asked of them?
Paul Patterson: Well, I think as Mr Beer asked me earlier, he had this concept of expert witness and witness of fact. It’s very serious, and these were dealing with serious matters. I did not feel, and clearly Ms Chambers didn’t feel the same, that it’s just a conversation, and I think it needed proper representation and engagement.
Ms Dobbin: Thank you. You also said, I think, that you were horrified, or certainly taken aback, that she was being contacted by a solicitor in order to deal with disclosure as well?
Paul Patterson: Yes.
Ms Dobbin: Again, is that for the same reason: that it’s really the company that should be involved in that process?
Paul Patterson: Well, whether it’s the company or not, I think Ms Dobbin, my point is that these are such serious matters for the case in front of the court and the impact, potentially, on a subpostmaster, it needed to be dealt with with the right professionals in the room taking the right advice and looking at the conversation of what was being asked of her in that particular case.
Ms Dobbin: Yes. To make sure it was understood –
Paul Patterson: Absolutely.
Ms Dobbin: – that what she needed to do.
I’m grateful to you, Mr Patterson. Thank you.
Sir Wyn Williams: Before anybody else asks a question, can I just be clear about one of the answers you gave to Ms Dobbin about the form upon which data, ARQ data, is sought by the Post Office. Am I to understand that that form is created by the Post Office without any input by Fujitsu or is it a combined effort, or what? I’m not quite sure where we are with that so I’d just like you to tell me again, if you could.
Paul Patterson: Sir Wyn, I think the way I understood the question was the form that is sent to Fujitsu requesting ARQ data – Mr Beer showed one, Ms Dobbin I think was referring to another version of it – that is a document that would have come from the Post Office to Fujitsu, I understand, saying what criteria to apply. I think it had evolved over time but it’s a document from the Post Office to Fujitsu.
Sir Wyn Williams: I follow the trail, so to speak. It goes from the Post Office to Fujitsu. My query is: who created that document in the first place?
Paul Patterson: Oh, I don’t know, sir.
Sir Wyn Williams: You don’t know. All right, thank you.
Mr Beer: Sir, before Ms Page asks her questions, I think I can help on that on a limited extent, although it’s slightly out of order, because it’s an issue that has raised its head now by virtue of your question.
I wonder whether we could look at FUJ00001318. That’s the wrong reference. This is what happens when you ask a question –
Sir Wyn Williams: Well, I’ll try to be quiet from now on, Mr Beer!
Mr Beer: No, I am going to retrieve the position. FUJ00152209. This was the Fujitsu 2005 document.
Sir Wyn Williams: Yes.
Mr Beer: It was a procedure or a policy document, interestingly dated – we’ve seen some correspondence about this in the course of today – 29 February 2005. 2005 wasn’t a leap year so it must have been misdated but, that aside, it’s excited lots of attention, we can find an annex within this document –
Sir Wyn Williams: Right, I see.
Mr Beer: – at page 28. Just like there was what I’ve called a boilerplate witness statement, a template witness statement, there was a template ARQ form earlier in the policy document. It set out, when you’re making an ARQ – or when the Post Office is making an ARQ request, this is the document that it should use.
Sir Wyn Williams: Right, so, effectively, it’s a Fujitsu created document which the Post Office then sends to Fujitsu to make its request?
Mr Beer: Yes. The position, as Ms Dobbin has said, does move on from this and we can explore with other witnesses and by looking at other documents how the amendments to this form were made, by whom and when.
Sir Wyn Williams: But that’s how it started, as a Fujitsu created document?
Mr Beer: Yes.
Sir Wyn Williams: Fine. Thanks.
Mr Beer: Thank you, sir, I hope that helps.
Sir Wyn Williams: Yes, it does, thank you.
Mr Beer: Ms Page, I think, was the last person who wished to ask questions.
Sir Wyn Williams: Over to you, Ms Page.
Questioned by Ms Page
Ms Page: Thank you, sir.
Mr Patterson, I appear for a number of postmasters, including Mr Castleton and Mrs Misra. First of all, the Inquiry has heard evidence that providing ARQ data to the Post Office brought in £850,000 per annum, which is obviously small beer in the grand scheme of things but, nevertheless, would you agree a reasonably handy bit of extra income?
Paul Patterson: I think that’s a very material sum of money.
Ms Page: So then there’s that and then, on top of that, there were charges for witnesses to appear in court and provide their evidence. We understand, for example, that in the trial of Tracy Felstead there was a charge of £20,000 for a witness to attend court. It was never, in fact, paid but that was the proposed sum of money. Have you ever looked into or found out about how much money Fujitsu charged for the services of Gareth Jenkins, for example, as a witness over the years?
Paul Patterson: No, I have not.
Ms Page: Would you accept that Litigation Support was a useful cash cow for Fujitsu?
Paul Patterson: I was professionally very surprised that that service even existed. We’re meant to be an IT company not a prosecution support service and, for that to be designed in from the very earliest stages, I was very, very surprised at it. And in terms of the work associated with doing it, I have no view on it. I am amazed that it was even in the contract.
Ms Page: Are you saying, in effect, that this is not something that Fujitsu was set up to do and not really something that Fujitsu was skilled at or able to do properly?
Paul Patterson: So if I look at today, Ms Page, there is no contract that I’ve never signed where that obligation and the extent of that obligation sits inside it. I do not understand why it was there. And, clearly, it was there for – and is still there, actually – for a very, very long time.
Ms Page: As a result of that, and moving on to my next question, we see the consequence, don’t we, of that sort of exceptional, and perhaps unprepared for, contractual service because, as we know, Anne Chambers gave evidence in less than ideal circumstances in the Castleton trial.
Paul Patterson: I agree.
Ms Page: In that trial, Mr Justice Havery ruled that Horizon did not cause the losses at Mr Castleton’s branch and that the Horizon shortfalls were “real deficiencies” caused by mismanagement at Mr Castleton’s branch. That was, in essence, founded on Ms Chambers’s evidence.
Do you agree that this blurring of the lines between witness of fact and expert witness on the overall system’s health was what led to this wrongful judgment?
Paul Patterson: I don’t know whether that led to it. As I said earlier in my answers to Mr Beer, I have learned, in preparation for this Inquiry, some of the differences between those two roles, an expert and a matter of fact. I do not believe, and Anne Chambers made it clear, that that was understood by the Fujitsu employee either. I don’t know whether it directly led to that, I’m afraid, Ms Page. I don’t have – I’m not skilled to know that.
Ms Page: Would you accept that it must have contributed materially to it?
Paul Patterson: Well, I think all of that would have contributed materially, and Ms Chambers was very clear – Mrs Chambers was very clear about her view of that.
Ms Page: Similarly, as part of that trial in the proceedings in the run-up to it, one of the formal court statements of case said “Fujitsu Services have looked at the Post Office’s computer system and have confirmed that the losses were not caused by the Post Office’s system’s software or hardware”, and the Post Office solicitor signed off on that with a statement of truth.
So, in other words, it was testifying to Fujitsu examining the system and finding that the losses were not caused by the system; do you know who at Fujitsu would have/could have authorised that verbiage?
Paul Patterson: No, I don’t.
Ms Page: Will you find out?
Paul Patterson: Well, I have – we have representation here and, if we can find out, we will absolutely find out. As to my earlier to the commitment to the Inquiry, we are delving into all the areas of information that we’ve got and, if it exists, we will find it. I’m afraid I don’t know at this present time the answer to your question, Ms Page.
Ms Page: Turning, then, to the trial of Seema Misra, in your statement you describe the duplicate data issue that affected the ARQ data that was served in that trial. As you say, Gareth Jenkins did actually provide a statement about it, didn’t he, and he let the Seema Misra defence team know about that one issue.
Paul Patterson: I believe that is in the exhibits here as well and in my – in the evidence in the third corporate statement, yes.
Ms Page: Would you agree that, as a one-off incident, that wasn’t perhaps the most serious of the data integrity problems but, as part of a series of data integrity problems, it was more concerning?
Paul Patterson: I think all the data integrity problems are serious and, earlier on today, we talked about the witness evidence and the clauses in bringing all of those points to disclosure for the subpostmaster and their representatives. So I think all of those things, whether it’s one item or the other four or five that we had discussed with Mr Beer.
Ms Page: Well, what I’m touching on here is the fact that one particular problem was disclosed and wrapped up in a bow, as it were, “Here it is, here’s the problem but we’ve fixed it”, and that that, in effect, sanitised the situation because this was really one of a series of problems with the audit data?
Paul Patterson: So I’d agree with that. I think, as we discussed earlier in evidence, the disclosure of all the things that could or could not impact needed to be disclosed and it is very evident in the witness statements that that was edited out.
Ms Page: Therefore, Seema Misra was not given the whole truth about the audit data problems?
Paul Patterson: I think that’s very clear.
Ms Page: Do you think that this came about as the result of pressure from above, on the likes of Mr Jenkins?
Paul Patterson: I really don’t know, Ms Page, and I’ve got no evidence before me that’s been produced for me in my submission today to the Inquiry that there was that, and I believe in other phases there’s more inquiry around governance, and such matters.
Ms Page: What has been done within Fujitsu to independently verify the investigation and remediation of these various audit data problems?
Paul Patterson: Just say –
Ms Page: Well, has there been any external or independent effort to investigate and remediate these problems?
Paul Patterson: I believe the current version of HNG-X or HNG, which is HNG Everywhere, is substantially different to the previous versions. I am unaware whether we’ve taken any third-party advice to go and audit that system again and the data that’s in it and how it’s discovered, Ms Page. I would need to look – I would need to take that one away, as a question for the Inquiry.
Ms Page: Thank you. Well, that would be helpful. Really, perhaps broaden that out a little, has there ever been any expert external investigation of the audit data function?
Paul Patterson: I am unaware whether there has been.
Ms Page: On a similar issue, has anyone been held internally accountable for these audit data problems at any stage, historically or more recently?
Paul Patterson: I am unaware of whether anybody has been and I think, in our letter to the Select Committee back in 2020, we undertook that if there were current employees who had done some of these things in a way that was against our policy, we would take disciplinary action against those individuals. I am unaware, Ms Page, whether we have found anybody along those lines.
Ms Page: Going back more historically, then, you will know about the problems with the EPOSS code, which were brought to light by David McDonnell.
Paul Patterson: I am – you’d need to give me more to go on, I’m afraid.
Ms Page: This is going right back to the start, that there were significant problems found with the EPOSS code and that they were, nevertheless – it was a decision taken not to rewrite and to allow the code to be monitored and watched in operation in the live state. Is that not ringing bells for you?
Paul Patterson: I remember a little bit on EPOSS, forgive me. I don’t know Mr McDonnell or the time frames around that, off the top of my head.
Ms Page: I think that probably answers my question, which, it seems, anyway, that nobody in current leadership has made the link or considered the link between the failures in the EPOSS code and the failures in the audit data that was being supplied through the Litigation Support Service. Nobody in the current team has made that link; is that a fair assumption?
Paul Patterson: I really don’t know. I know what I know and I don’t know the particulars of that.
Ms Page: All right.
Paul Patterson: We can – certainly the team, we will take that one away.
Ms Page: Well, again, that sounds very helpful, Mr Patterson, and certainly another – again, another question arising from that is the historic question: did nobody historically make that pretty obvious connection between very poor code going out into operation and then very poor data coming out through the Litigation Support Service?
Paul Patterson: Whether people made that connection or not, what is very evident, I think in our evidence today and in the previous corporate statements, is that that connection and understanding about what was going on and where was it was understood by certainly Fujitsu and certainly understood by Post Office, right back to 1999. It’s all about what you do with that information and I think that’s the key – forgive me, that is a question for this Inquiry, I know that you – that the Inquiry is testing.
Ms Page: Yes. Thank you.
Thank you, sir.
Sir Wyn Williams: I think I agree with you, Mr Patterson: it is a key question.
Right, any other advocates wish to ask any questions?
Mr Beer: No, sir, there are not.
Sir Wyn Williams: Well, thank you very much, Mr Patterson, for coming to give oral evidence and, obviously, at least participating in making three very detailed witness statements. I think you understand that you will, in all probability, be recalled at a future date but, for the time being – well, not for the time being – you are now released, so to speak, from your status as a witness, which means that you can continue having discussions with your team, et cetera, to answer any questions which the Inquiry asks of you in the future and to help you make any further witness statements that are necessary. All right?
Normally, when a witness is to give further evidence, there is an embargo on such discussions but that’s not practical in an Inquiry of this sort, so, as I’ve said, you’re free to speak to your team about all relevant matters now.
The Witness: Thank you.
Sir Wyn Williams: All right.
So we’re ready to start at 10.00 on Tuesday, Mr Beer, is that it?
Mr Beer: Yes, that’s right, sir.
Sir Wyn Williams: All right. See you then. Bye.
Mr Beer: Thank you very much.
(2.40 pm)
(The hearing adjourned until 10.00 am on Tuesday, 23rd January, 2024)