1 Tuesday, 5 September 2023 2 (10.00 am) 3 MR BEER: Sir, good morning, can you see and hear 4 me? 5 SIR WYN WILLIAMS: Yes, I can, thank you. 6 MR BEER: Thank you very much. As you know, you 7 directed this hearing today to investigate 8 a series of disclosure failings by the Post 9 Office and also to hear evidence about what has 10 been done to put those disclosure failings 11 right. The Inquiry asked for five witnesses to 12 produce written evidence and duly received five 13 witness statements from those witnesses. 14 You'll hear evidence from four of those 15 today: Fintan Canavan, who Mr Blake will be 16 asking questions of firstly; Diane Wills, who 17 I will be asking questions of; Paul Tombleson, 18 who Mr Blake will ask questions of; and then 19 finally Gregg Rowan, who I will examine. 20 The fifth witness was Mr Ben Foat from whom 21 you heard previously, the Post Office's general 22 counsel. The Inquiry took the view, as a result 23 of reviewing the contents of his witness 24 statement, that more meaningful and detailed 25 evidence could be obtained from the other four 1 1 witnesses that I have mentioned and, therefore, 2 he is not to be called today to give oral 3 evidence. His witness statement will be 4 uploaded, as will the other four witness 5 statements, to the Inquiry website today. 6 So, sir, can we start then, please, with 7 Mr Blake and Mr Canavan. 8 SIR WYN WILLIAMS: Certainly, but can I just make 9 sure that I'm up to date with the documentation. 10 I received this morning a letter or an email 11 from Mr Henry, which he sent in the middle of 12 the night, and I received further information in 13 a letter from the Post Office which arrived at 14 8.20 this morning. Am I up to date, Mr Beer? 15 MR BEER: Yes, you are, sir. The 3.10 am email from 16 Mr Henry was itself a response to a 22-page 17 letter received from the Post Office yesterday, 18 which the Inquiry distributed. In summary, 19 Mr Henry says it's difficult to digest complex 20 22-page letters on the eve of a hearing. Please 21 can you impose a deadline in future that the 22 Post Office should, if they wish to update the 23 Inquiry with information, do so not less than 24 48 hours before a hearing. 25 Then the 8.20 am communication this morning 2 1 is an amended version of a table that was 2 disclosed at 4.26 pm yesterday. You are up to 3 date, sir. 4 SIR WYN WILLIAMS: Yes, thank you. All right then, 5 over to Mr Blake. 6 MR BEER: Thank you. 7 MR BLAKE: Thank you, sir. Can I call Mr Canavan, 8 please. 9 FINTAN CANAVAN (affirmed) 10 Questioned by MR BLAKE 11 MR BLAKE: Thank you very much, can you give your 12 full name please? 13 A. Fintan John Canavan. 14 Q. Mr Canavan, you should have in front of 15 a witness statement dated 22 August 2023; is 16 that correct? 17 A. It is. 18 Q. If you turn to the final page of that statement 19 can you confirm that that is your signature? 20 It's on page 17 of 17. 21 A. It is. 22 Q. Is that statement true to the best of your 23 knowledge and belief? 24 A. It is. 25 Q. Thank you very much, Mr Canavan that witness 3 1 statement will go on to the Inquiry's website. 2 It's unique reference number is WITN09970100. 3 It doesn't need to be brought on to the screen 4 just yet. I'm going to start just by asking you 5 a brief career history. You're a qualified 6 solicitor and I think you've been practising for 7 over 30 years; is that right? 8 A. That's correct, I was admitted into the Law 9 Society of England and Wales. I then 10 transferred and I'm admitted to the Law Society 11 of Northern Ireland. 12 Q. You've acted for Core Participants in a number 13 of public inquiries including, for example, the 14 Bloody Sunday Inquiry? 15 A. That's correct. 16 Q. You're currently a partner at DAC Beachcroft 17 solicitors and have been in that position since 18 September 2020; is that right? 19 A. That's correct. 20 Q. In July 2021 you were asked by a colleague 21 whether you'd be willing to be seconded to the 22 Post Office to assist with their internal 23 Inquiry Team; is that correct? 24 A. That's correct. 25 Q. Were you aware at that stage, so in July 2021, 4 1 that on 1 June 2021, the previous month, the 2 Terms of Reference for this Inquiry had been 3 announced? 4 A. No. 5 Q. You were eventually contacted by the Post Office 6 in October; is that correct? 7 A. It was around October, yes. 8 Q. Were you aware that, over the course of the 9 summer before that October, before you were 10 contacted, the Post Office and 218 others had 11 become Core Participants in this Inquiry? 12 A. No, I had not followed the Inquiry at all until 13 I actually became involved with it. 14 Q. When you were contacted in October, were you 15 briefed on that kind of background? So, for 16 example, by October a provisional list of issues 17 had already been published, including some 184 18 issues? 19 A. No, I had a number of phone calls with parties 20 from the Post Office and it was much more about 21 my own background, what my understanding of 22 inquiries was, what my experience with inquiries 23 had been and to indicate that the Post Office 24 needed some assistance with an internal team to 25 assist the Inquiry. But we hadn't got into any 5 1 discussions about what the extent of the Inquiry 2 itself was at that initial point. 3 Q. Did you know why it wasn't until October 2021, 4 so four months after the Terms of Reference had 5 been published, that you were contacted? 6 A. No, it was -- I'd been contacted in July. 7 I think one of my colleagues in London who works 8 with -- had done some work with the Post Office 9 had indicated to me that there was a desire to 10 bring somebody in. I indicated I would be happy 11 to do so and then it was a number of months 12 later when the actual contact to see if I would 13 be available took place. So no, there was no 14 discussion about what had taken place during 15 that period. 16 Q. How long after the phone call did you actually 17 join the Post Office team? 18 A. Probably within about a week or so. It was -- 19 there were a number of calls, one initially with 20 the -- Mr Mark Underwood, who was the compliance 21 director, I can't remember his exact title but 22 he worked in the compliance team -- and then 23 a second call with Mr Foat. 24 Q. Your title was Inquiry Director; is that 25 correct? 6 1 A. Correct. 2 Q. You've spoken about Mr Foat. You reported 3 directly to Mr Foat, who was the general 4 counsel; is that right? 5 A. Correct. 6 Q. Although you are a lawyer, you weren't actually 7 acting in a legal role; is that right? 8 A. That is correct. It was quite clear. 9 I wasn't -- I haven't -- I wasn't SRA registered 10 at the time. I am admitted to the role but 11 I wasn't SRA registered and the discussion was 12 clear that my role was the Inquiry Director; the 13 legal advice and legal representation was by the 14 external firm, who were Herbert Smith Freehills. 15 Q. Can you briefly talk us through where you sat 16 within the company in respect of, say, the group 17 executive, various steering committees and the 18 external legal firms. 19 A. Without trying to give a very detailed 20 description of it, the General Executive -- 21 Mr Foat was a member of the General Executive 22 and I reported directly to him. So we were -- 23 the Inquiry Team was based in Mr Foat's area, 24 the sort of the legal and governance side of the 25 business. I chaired the Inquiry SteerCo, which 7 1 initially met every week, or -- it would have 2 met as much as we needed it to meet but it was 3 listed for a hearing or a meeting every week and 4 on the steering committee were a number of GE 5 members. 6 It was chaired by me and then was attended 7 by whoever needed to attend. So there would 8 have been Inquiry Team members; HSF would have 9 attended on a number of occasions; Peters & 10 Peters. As and when an issue arose, the 11 relevant people were invited to attend. That 12 would have sat regularly. 13 I would also have sat in the Freedom of 14 Information, the FOIA SteerCo but my role on 15 that was a secondee into that role. I didn't 16 have any voting or quorum rights within that 17 steering committee. My role was to ensure that 18 if we were aware of information that might be 19 relevant to a Freedom of Information request or 20 if a Freedom of Information request addressed 21 something that we would need to be aware of, 22 there was a link in that regard. 23 Q. You've mentioned two external legal firms: 24 Herbert Smith Freehills and Peters & Peters. 25 What do you see as the difference between their 8 1 two roles? 2 A. Herbert Smith Freehills were the lawyers 3 representing Post Office with the Inquiry. 4 Their role was to advise -- to link between the 5 Post Office and the Inquiry Team. They were our 6 advisers and representatives. Peters & Peters, 7 as I understood it, they dealt with a lot of the 8 prosecutions historically, they had a lot of 9 material available in regard to the prosecutions 10 and they would have assisted particularly in 11 regard to those -- that aspect and the phases 12 which would have dealt with prosecution 13 decisions. 14 Q. You've spoken about the various steering 15 committees, the Group Executive, et cetera. Was 16 there anyone other than yourself with 17 significant experience in public inquiry work? 18 A. Not as far as I'm aware. 19 Q. I want to ask you about the length of the 20 Inquiry. You've said in your statement it was 21 anticipated the Inquiry would only last for 22 three to four months. Who was it that told you 23 that? 24 A. That was the indication when I was contacted in 25 July to see -- I have a full -- a fairly 9 1 substantial caseload. My background, I do a lot 2 of Legacy litigation and public inquiry work, so 3 I had a fairly extensive caseload, and the 4 indication was to me it would only be three or 5 four months, so there would be no need for me to 6 try to offload my caseload or to try to make 7 arrangements within the practice. And then when 8 I spoke in October, I think the anticipation 9 still then was that it wouldn't be running for 10 an awful lot longer than that. 11 So initially when I came in, I did not make 12 any arrangements or alterations to my caseload 13 and I was still trying to manage that. 14 Q. Who was it within the Post Office that gave you 15 that impression? 16 A. My conversations were initially with 17 Mr Underwood and with Mr Foat and those initial 18 discussions were that it wouldn't be a long 19 period of secondment within the Post Office. 20 Q. Did you have a conversation then about retaining 21 your current caseload? 22 A. Only internally within the business that I had 23 to speak to my line management team to make sure 24 that there was sufficient cover if court 25 hearings, whatever, came up, that I would not be 10 1 able to attend those. But that discussion was 2 internal. It wasn't externally with the Post 3 Office. 4 Q. Given the experience that you have had -- for 5 example the Bloody Sunday Inquiry lasted 6 12 years -- did you think that that was 7 a realistic time frame? 8 A. Initially, I didn't question it because I hadn't 9 had any involvement, so I wasn't aware of what 10 conversations and discussions had gone on. 11 I wasn't aware of the extent and nature of the 12 matters under this Inquiry. When I first came 13 in, within a very short space of time, it was 14 quite clear that the team we had was too small, 15 that the expectation of time was too small and 16 that the budget we had was just too small, that 17 it was always going to expand and that even 18 initial expansion expectations were too short. 19 Q. You address this to some extent in your witness 20 statement. I'd just like that to be brought on 21 to screen. It's WITN09970100. Can we look at 22 the bottom of page 2, it's paragraph 8, please. 23 You say there: 24 "It was immediately apparent that the scope 25 of the Inquiry, the period under investigation 11 1 and the issues involved meant that the 2 anticipated timescale and the team within POL 3 was inadequate. This created an immediate issue 4 with the budget provided for this project 5 (a theme which underpinned much of the following 6 work in the following 15 months)." 7 You say there it was "immediately apparent"; 8 was that to you or to others as well? 9 A. With the conversations we were having, that was 10 one of the initial discussions that we would 11 have had, that this isn't going to work: (a) 12 three to four months was not going to be 13 sufficient. There was no way that -- bearing in 14 mind the Inquiry hadn't even started having 15 hearings we were not going to complete in 16 January/February of the following year. That 17 did become apparent very quickly and steps were 18 being taken very quickly at that point. 19 Q. Did it come as a surprise to those you were 20 having conversations with? 21 A. I don't think people within Post Office 22 understood the nature and I -- that's not 23 a criticism of the Post Office. I don't think 24 many people understand the nature of an inquiry 25 as opposed to a trial, in that a trial can be 12 1 listed months in advance for five days, maybe 2 goes to six or seven, but you know what your 3 timescales are. 4 Inquiries by their nature are much more 5 fluid and I don't think there is 6 an understanding of the nature of that fluidity 7 and the organic nature of an inquiry, which can 8 change direction simply because of information 9 which comes to light. So I don't think it was 10 understood and I think it did come as a surprise 11 to those within the management level of the Post 12 Office as to the extent that the Inquiry could 13 expand and how long it was going to take. 14 Q. Moving on to budget and I'll take you again back 15 to your witness statement, paragraph 13. It's 16 page 4. You say there: 17 "As the Inquiry developed and grew, the 18 issues around the projected budget and the need 19 to secure a higher budget became the core focus 20 and took up a significant part of the time 21 I spent in the role. A very significant amount 22 of my time was spent amending budgets, 23 forecasting and projecting different scenario 24 budgets, seeking approvals for increased budgets 25 and drawdown, discussing ways to reduce fees 13 1 with HSF and looking at ways to achieve the same 2 outcomes at lower cost. This meant less and 3 less of my time was spent on the actual 4 processes themselves." 5 Can we turn over to page 13 and in 6 paragraph 50, so later on in your statement, you 7 come back to the theme of budget. You're 8 addressing there the rectification of legacy 9 document storage and you say it: 10 "... does not feature highly in the 11 priorities of any area and there is a degree of 12 passing the buck hoping other areas will take up 13 the responsibility." 14 You say: 15 "Again this goes to the budget allocations 16 and the need for those heading each business 17 area or sub-area within those workstreams did 18 not allow for much leeway to allocate staff, 19 time or money to document retention, 20 identification and storage." 21 Further down in paragraph 52, you give 22 an example of the storage facility in Winchester 23 and you say: 24 "Much [of it] has not been accurately 25 indexed requiring manual searches of high 14 1 volumes of material. Boxes were found to 2 contain lottery receipts, memos, advertising 3 brochures all in one space and if time and money 4 were available much of the material could be 5 reviewed and destroyed but that is not 6 possible." 7 In your view, were there sufficient funds 8 available at the Post Office for a comprehensive 9 disclosure exercise? 10 A. I actually don't think that's a simple question 11 to answer, in that the Inquiry's budget came 12 from within the general Post Office budget. The 13 general Post Office budget was partially 14 controlled by the Government department, but 15 partly they were a private company who had to 16 generate a certain amount of their own income. 17 And projecting budgets, I don't think any 18 business could have anticipated the expenditure 19 that would have been needed to rectify 20 years/decades of document disclosure and I -- 21 I have a lot of sympathy for those within the 22 Post Office themselves who were expected to run 23 their departments, continue the operation of the 24 Post Office but then also find time and the 25 money out of limited budgets to assist me in 15 1 finding other documents. 2 I think, in short, no, there wasn't 3 sufficient money within the Post Office to both 4 run the business and deal with the legacy 5 aspects that were highlighted during the process 6 of the Inquiry. 7 Q. Mr Canavan, you talk about the process of the 8 Inquiry but there had, of course, before that 9 been Court of Appeal proceedings, there had been 10 Group Litigation, all of those required 11 disclosure to the courts. Looking back at the 12 funding that was in place for simply maintaining 13 and resourcing the document management within 14 the Post Office, do you think that there was 15 sufficient resourcing and, if not, why not? 16 A. No, I don't. I think that possibly the 17 consideration hadn't been given to the legacy of 18 material that existed and I think that example 19 of when we were required to go into Winchester 20 and we found thousands of boxes which hadn't 21 been indexed at all, other boxes were indexed 22 inaccurately, and when you started going through 23 it, the material -- and, in a way, the defects 24 have assisted this Inquiry. 25 Had there been a proper process of document 16 1 retention, disclosure, organisation and 2 destruction, many of these documents would have 3 been destroyed within a reasonable -- a policy 4 of destruction after, say, 10 years because 5 material was no longer used. A lot of that 6 material, had it been retained in a structured 7 way, may not have existed now. 8 But I do think the issue of document 9 identification, location and storage is not 10 a priority, and it may well be now, because of 11 the issues which have arisen but I don't think 12 it was and I think the issues that we 13 identified, with boxes being found in different 14 offices and different storage areas, highlighted 15 that there wasn't an understanding within the 16 business as to where all of their material was 17 stored. 18 Q. In your view, who was responsible for that? 19 A. I think it's a legacy thing. To say who was 20 responsible, you would need to go back to when 21 the material was being initially stored, so go 22 back to the Royal Mail period where someone is 23 in the Postal Museum, someone is in Winchester, 24 the material we found in the Londonderry Crown 25 Office, which as I said, from my memory, when we 17 1 had the safes cracked open, predated Horizon. 2 So that was material that people didn't even 3 know they had. 4 So I don't think you can blame or allocate 5 responsibility to a small number of people. 6 I think it is a legacy across all of the running 7 of the business. But I think ultimately the 8 General Executive should have an understanding 9 of what material is in their business and I say 10 that in a global context, not as a criticism of 11 any member of the current General Executive who 12 have inherited this issue, as opposed to created 13 it. 14 Q. Looking at, for example, the corporate 15 recordkeeping, in broad terms, as 16 an organisation, can you describe the state of 17 the repositories that you found? 18 A. I have no personal involvement. I didn't 19 actually attend any of the sites. The reports 20 that were coming in to us would suggest that 21 there wasn't a single repository of any 22 particular material. So it was scattergun in 23 its storage. I think the core corporate 24 material, so the board meeting minutes, we 25 did -- from memory, I think we were able to find 18 1 those fairly easily, in that they were -- the 2 older ones were in the Postal Museum and many of 3 the others were then locatable within the 4 company secretariat. 5 Q. You've described in your witness statement 6 "fractured and disjointed". Can you elaborate 7 on that a little bit? 8 A. Exactly what I'm saying where you have some 9 material in the Postal Museum, you have some 10 material in Winchester, you have some material 11 that was in Finsbury Dials, you had material 12 held in large regional Post Office hubs, and 13 there wasn't a single point of contact where you 14 could contact one person and say, "I'm looking 15 for A, B and C", and they'd be able to say, 16 "That's stored in this location". 17 It did involve them having to conduct very 18 widespread searches physically and 19 electronically to try to locate material. 20 Q. This Inquiry had been on a non-statutory footing 21 for quite some time and then, in June 2021, the 22 Terms of Reference were announced as a Statutory 23 Inquiry. You joined in October of that year. 24 When you joined, were efforts already in place 25 to try to improve the situation or were you the 19 1 first to make the suggestion? 2 A. I think it was an organic process, where, as the 3 requests came in for material, the level of 4 disjointed storage became known and it wasn't 5 apparent immediately that from day one, you had 6 arrived -- and when I arrived, my first core 7 objectives were that the team was too small and 8 the budget was too small. So we weren't hit 9 immediately with a lot of the disclosure 10 requests. 11 I think maybe some of those came in 12 around -- I remember there were some around 13 December but I think at that point, even then, 14 we were more internally concerned with getting 15 the team in place and a lot of the disclosure 16 requests were dealt with externally. 17 Q. Were you aware, for example, that Mr Justice 18 Fraser in the Group Litigation had made 19 criticisms of Post Office's disclosure? 20 A. I had read the judgments, the two core Fraser 21 judgments, so I was aware that there was 22 criticism of that. 23 Q. Had those criticisms permeated throughout those 24 who you worked with at the Post Office? 25 A. Yes, it was -- the issue of disclosure was very 20 1 prevalent, it was of high priority within those 2 that I dealt with. I think I raised the point 3 in my statement where discussions took place 4 about how did we ensure that there was no issue 5 on disclosure, and there were suggestions of 6 should we not just hand the keys to Relativity 7 over to the Inquiry so they had access to 8 everything or can we not just -- if they want 9 boxes from Winchester, can we not just bring the 10 boxes, so there was no suggestion of anything 11 being hidden, and that was not appropriate. 12 It's not appropriate because the Inquiry 13 does not have the resource to go through that 14 amount of material. But, likewise, there would 15 be a concern that there was an attempt to hide 16 information, by simply blizzarding the Inquiry 17 with disclosure that they would miss material. 18 So it was understood that there was a duty 19 on us to provide everything the Inquiry wanted 20 to see and there was a clear desire within the 21 GE members that I spoke with that we did need to 22 make sure that you had everything you wanted. 23 Q. Given that there was that desire, why was it 24 that it wasn't sufficiently funded? 25 A. I think you can only make money go so far. The 21 1 Post Office's budget was, as I understood, there 2 was a Government-set budget. The Post Office 3 has specific duties and responsibilities that 4 they have to comply with. They have to keep 5 a certain number of post offices open, they 6 still have to recruit and train postmasters, 7 they still have to replace the Horizon System. 8 So there was an element of prioritising and 9 there were efforts to move budgets and that 10 was -- a constant theme throughout my time was 11 how do we -- how much money does the Inquiry 12 need? How much money does the Inquiry Team 13 need? Where can we find the money? 14 And the message is going out to the BAU 15 departments that, if there were processes that 16 they could stop and save money they should do 17 so. But they only had a limited budget that 18 they could apply to the entire Inquiry process. 19 That budget had to cover not only the document 20 retention but the Inquiry Team, HSF, Peters & 21 Peters, maintaining the Relativity platform. 22 That all came under the same budget. So finding 23 extra money to do investigative processes, such 24 as going out to Winchester, re-indexing and 25 doing tours around all of the offices to locate, 22 1 had to be found from other resources and I think 2 that was just difficult to do. 3 Q. Diane Wills has said in her statement -- and 4 we'll hear from her shortly -- that the team had 5 been under-resourced for a long period of time. 6 Were you unsuccessful in your efforts to try and 7 get more resources? 8 A. I think a lot of work was done in my time, up to 9 the point when I left, where we were doing 10 different scenarios where, if we could increase 11 the Inquiry's internal team and reduce the 12 external cost, what would that be? Where would 13 we get the funding? And that debate about the 14 different ways and different processes was under 15 way. 16 I think maybe if Diane has increased her 17 funding since I left, then that work was 18 probably done before I left, and it was clear 19 that steps had to be taken to increase the 20 funding. I think we had regular meetings and we 21 had debates of budget forecasts that went to the 22 Department. We had to go to the board. The 23 board themselves were generally very supportive, 24 when we needed money to draw down the board, the 25 GE were supportive of that, but it was within 23 1 the budgets that they had available. 2 Q. I want to ask you about the role of Herbert 3 Smith Freehills. You've addressed that in your 4 statement at paragraph 25 onwards. You say that 5 the instruction predated your role at the Post 6 Office but you say you believe it was influenced 7 by their previous involvement in Post Office 8 related litigation and also the anticipated 9 short duration of the Inquiry. I want to take 10 those one by one. 11 So taking the first of those, what did you 12 believe the relevance of Herbert Smith's 13 previous involvement to be in their instruction 14 in the Inquiry? 15 A. When I arrived, they had -- they were already in 16 place and, as I understood, the process earlier 17 in the year of appointing external lawyers had 18 been influenced by the already present position 19 within Post Office regarding the legacy 20 matters -- excuse me -- that Herbert Smith 21 Freehills -- I believe they were involved in the 22 compensation process, that they had already 23 become involved in a compensation process and, 24 from memory, I haven't gone back to it, but they 25 may have been involved in the 555 process -- 24 1 So they had some deeper understanding of 2 some of the issues that were going to come up in 3 the Inquiry and, as I understood it then, the 4 decision at senior level was it made sense to 5 utilise the information which was already 6 available to Herbert Smith Freehills, than to 7 bring in a new firm who would then have to come 8 up to speed with all of the issues that they had 9 already become involved with. 10 Q. Were you aware, for example, that the Group 11 Litigation was going to be something that was 12 going to be investigated by the Inquiry? 13 A. When I started, no. But it was a clear issue 14 that would obviously be of relevance to the 15 Inquiry to understand the previous trials, 16 although when I first started, I think the 17 understanding was that the Inquiry would not be 18 seeking to go back into the previous litigation. 19 So I may not then have appreciated that taking 20 the Fraser trials from one side and the Group 21 Litigation as a separate process, that they 22 would all have been in my head one earlier 23 process, and it did make sense that Herbert 24 Smith Freehills had that prior knowledge. 25 Q. In terms of the second factor, why would the 25 1 length of the Inquiry be relevant to Herbert 2 Smith's instruction? 3 A. As became clear, the budgetary aspect of the 4 Inquiry, in that I don't believe there was any 5 negotiation of a budget or a fee process with 6 Herbert Smith before I arrived, that was already 7 in place by the time I took up my role. So 8 a shorter duration would not have had 9 a significant budgetary impact on the wider Post 10 Office. But a longer duration, it would have 11 made more sense to look at other alternatives 12 that may have been less expensive or perhaps 13 some kind of fee negotiation with Herbert Smiths 14 that may have taken place. 15 Q. Is this because, as you've previously described, 16 there was this fixed pot of money that wasn't 17 expanding? 18 A. I think the Post Office's budget is fixed across 19 its entire -- and then the internal allocation 20 of that budget. There had been a small budget 21 allocated to the Inquiry process which did 22 continue to expand but I think the shorter the 23 duration, the less relevant the budgetary 24 implications would have been. The longer the 25 process, the more relevant budgetary 26 1 negotiations would have been to any process. 2 Q. Was anyone at this stage raising it with the 3 Group Executive or somebody senior within the 4 Post Office that they urgently needed more 5 funding? 6 A. Well, it was apparent and I had conversations 7 with Mr Foat, who was frustrated, and his 8 approach was very much "We need to get this 9 done, we will have to find the money". But 10 again, that's easy to say, "We will just have to 11 find the money". The money then has to come 12 from somewhere. So the issue of funding was 13 a constant theme and, whilst it may not have 14 been openly dealt with, it was a constant 15 process of "How much will this cost, what do we 16 have to spend to do that?" 17 And one of the points that I think I did 18 raise was when HSF would have said to us "We 19 need to send five people to Winchester for five 20 days", they would have come to SteerCo and said, 21 "This is going to cost roughly this much", that 22 would then have been approved. And I'm not 23 aware, I don't recall any instance when that 24 expenditure was ever declined. It was always 25 approved. But monitoring and watching that 27 1 budget was essential because there was a limited 2 budget. 3 Q. You say at paragraph 26 of your witness 4 statement that the approach to each disclosure 5 request was designed by Herbert Smith Freehills. 6 Perhaps we can look at paragraph 53, it's 7 page 14. Thank you. You say at paragraph 53: 8 "HSF sought to identify relevant search 9 criteria based on each R9 request. Those terms 10 were devised by and applied by HSF and were not 11 discussed with POL." 12 Looking back, do you think that was 13 an appropriate action to take? 14 A. Yes, I do. The team within Post Office would 15 not have understood what the -- the search 16 criteria were set up by the external advisers, 17 who were immersed in the process. They would 18 have set the criteria to what the Rule 9 Request 19 was looking for. Some Rule 9 Requests were very 20 simple, you could identify a specific document. 21 That was easy. But if it was wider requests, we 22 relied on HSF to tell us they have received 23 a Rule 9 Request, it relates to -- and they will 24 have disclosed the Rule 9 Request to myself, but 25 they would have devised then what they needed, 28 1 what they felt was the appropriate search, and 2 they did so under the direction that it was to 3 be as wide as possible and it wasn't to be 4 a restrictive search. 5 Q. But in terms of providing assistance to the 6 Inquiry, getting the right documents that the 7 document wanted, do you think it is appropriate 8 for an external law firm to devise and apply 9 search terms and not to discuss them with their 10 client? 11 A. When I say they weren't discussed, they wouldn't 12 have sent a note to us saying, "We are applying, 13 'Horizon', 'Horizon litigation'". They didn't 14 necessarily set them out. They would have -- 15 each Rule 9 Request would have been brought up 16 at a SteerCo, they would have identified where 17 they had identified relevant areas that they 18 needed to search, be that Postal Museum, 19 Winchester or Relativity searches. They would 20 have indicated to us that -- we received -- 21 I think sometimes there were hundreds of 22 thousands of hits that went to first level 23 review, that was reduced then to a second level 24 review and then the disclosures were then sent 25 to the Inquiry when the relevant material was 29 1 identified. 2 Q. Did they too had on every occasion? 3 A. My memory is that yes, they did. I don't 4 believe there were any requests where we were 5 not aware of the Rule 9 or the nature of the 6 searches, primarily because any budget 7 requirement to send people out to the hard copy 8 repositories would have had to have been 9 approved. 10 Q. In terms of human involvement, though, from the 11 Post Office, in terms of somebody with 12 experience, with knowledge of Post Office 13 policies, procedures, it sounds from paragraph 14 53 of your statement that there wasn't that kind 15 of involvement in, for example, search terms? 16 A. It may well have been -- I assume that may be 17 slightly misleading in that they did not discuss 18 the phrasing of the search terms or how they 19 would have framed and carried them out. The 20 Inquiry Team was kind of two layered. There was 21 the legal side where we had legal counsel who 22 would have been involved with certain work, but 23 there was also an operations side, and there 24 were members on the operations side who were 25 longstanding Post Office employees, and they 30 1 would have had contact with HSF and HSF would 2 have utilised that knowledge about questions -- 3 they could have said "Have you looked at", or 4 "I remember that incident", or "I remember 5 a person who was involved in that department". 6 So there was contact -- 7 Q. So although you've said in your witness 8 statement there those term were devised by and 9 applied by HSF and were not discussed with Post 10 Office Limited, your evidence is that there were 11 people within Post Office Limited, in the 12 operational side, that did have discussions with 13 Herbert Smith Freehills? 14 A. They -- the discussions would have been about 15 the process, not the terms. Herbert Smith 16 Freehills devised what searches were required, 17 what the -- what the Rule 9 was requesting, how 18 to conduct those searches, what they would be 19 looking for, but they would have had access to 20 anyone within the Post Office had they needed 21 access to anyone in particular for advice on 22 where to go looking for material. There would 23 have been no -- Post Office did not dictate or 24 direct what the search terms were or what the 25 search criteria should be. 31 1 Q. So you have said that they had access to people 2 at the Post Office? 3 A. Yes. 4 Q. Did they routinely consult people in the Post 5 Office with regards to requests that had been 6 made? 7 A. There was consistent two-way conversations going 8 on that the contact would have been made through 9 both SteerCo and through requests for 10 information and any information which we devised 11 or located. So if one of our processes located 12 material, we would have immediately notified HSF 13 and if they had needed access or if they had 14 requested any access, it would have been 15 granted. And I do recall instances when people 16 within the operations team did speak with HSF. 17 I couldn't give you information as to how 18 regular or consistent that was. 19 Q. Was there a policy in place as to in what 20 circumstances to speak to somebody at the Post 21 Office? 22 A. No, there was no formal policy. 23 Q. Perhaps we can take an example, and that's the 24 request for prosecution policies and 25 investigation policies. Can we look at 32 1 INQ00002007, please. This is 28 February, 2022, 2 Rule 9 Request. It's Rule 9(11). This was 3 a request while you were in post and can we look 4 at paragraph 15 of this request. Thank you. 5 This request says: 6 "The Minutes of the Audit, Risk and 7 Compliance Subcommittee of 11 February 2014 8 refer to a report which outlined the proposed 9 changes to the prosecutions policy and a paper 10 to explain the most appropriate way to 11 communicate the prosecutions policy. Please 12 provide copies of the same and copies of all 13 iterations of the prosecutions policy since 1999 14 that are in POL's custody or control." 15 If we move down the same request 16 paragraph 46, please. Sorry, it's a bit above 17 that. Thank you. There's a reference there to 18 the "Minutes of the Audit, Risk and Compliance 19 Committee", and it says: 20 "These refer to an investigations policy 21 that was circulated. Please provide copies of 22 the same and copies of all iterations of the 23 investigations policy since 1999 that are in 24 POL's custody or control." 25 There was a later Rule 9 asking for much the 33 1 same. That's INQ00002008. So that was 2 a request of 15 June 2022. Perhaps we can look 3 at paragraph 18. That request asked for: 4 "Policies and guidelines ... relating to the 5 bringing of private prosecutions against 6 subpostmasters and other end users ..." 7 Request 22 says: 8 "Any guidance, training or instruction ... 9 given to those responsible for conducting 10 criminal investigations ..." 11 Paragraph 26, scrolling down, similarly asks 12 for: 13 "Any guidance, training or instruction given 14 to those responsible for making charging 15 decisions ..." 16 Paragraph 30: 17 "Any guidance, training or instruction given 18 to those responsible for conducting the 19 prosecution of subpostmasters ..." 20 Just by way of one more example, paragraph 21 34: 22 "Any guidance, training or instruction given 23 to those responsible for disclosure in criminal 24 proceedings brought against subpostmasters ..." 25 These requests were made while you were in 34 1 post. The Inquiry's understanding was that the 2 process that followed that was that Herbert 3 Smith Freehills would set search terms, the 4 repositories would be searched and the end 5 product would come back. We know that 6 a significant document, the racial profiling 7 document, was not contained in what was 8 ultimately produced. 9 Is your evidence then that there was 10 a discussion about prosecution policies with 11 somebody in the Post Office, between them and 12 Herbert Smith Freehills? 13 A. I can't answer that question in that I don't 14 have any specific memory of this individual 15 request. My understanding would be that Herbert 16 Smith would contact Post Office identifying 17 particular search areas, so the hard copy/soft 18 copy repositories. If there was a specific 19 issue, in this case looking at governance 20 policies and procedures, that would have been 21 referred to the Inquiry team. That, I would 22 believe, would have been referred to the 23 Secretariat, the company secretaries department, 24 and any responses would have been identified, 25 passed to Herbert Smith, and then passed on. 35 1 Q. Your evidence is that it would have happened? 2 A. That would be my belief. 3 Q. How involved in this process were you as Inquiry 4 Director? 5 A. I was -- I had very little direct involvement in 6 any of the specific processes, particularly by 7 the time we came to June of last year, we were 8 heavily involved in trying to expand, even in 9 June last year, the process had been that we 10 would have been ending the Inquiry by sort of 11 Christmas and then it was quite clear it wasn't 12 going to end, and budgets were severely 13 constrained by that point. 14 The process, I think -- maybe when I came 15 into post it was a very small team, we expanded 16 the team both on the legal and operational side 17 and, just before I departed, an Operations 18 Director had been appointed. The Head of Legal 19 who had been on maternity leave was due back and 20 my understanding would have been that I would 21 have then moved into the role of overseeing 22 rather than being hands on, day to day, dealing 23 with the finance team, and drafting of the 24 processes. I would have been a much more 25 oversight role. 36 1 The majority of the requests would have been 2 dealt with by the individual teams that were 3 created. The process we had, Phases 2 and 3 4 were allocated to two counsel, Phases 3 and 4 5 were allocated to two counsel, 5 and 6 were -- 6 is it -- yeah, were allocated then to 7 a different set of counsel, and then they would 8 have been tasked to deal with their counterpart 9 in HSF about any requests that were relevant to 10 their area, that they would have assisted in 11 that process. 12 I would not have had any day-to-day or even 13 detailed responses or updates on specific 14 documents that were located. 15 I think had the racial profiling document 16 been identified, I would have been aware of it 17 and I would have been made aware of it because 18 of the gross nature of the material which had 19 been produced. 20 Q. So as Inquiry Director, your role was not to get 21 involved in the nitty-gritty of the Rule 9 22 requests? 23 A. No. 24 Q. That was left to legal counsel? 25 A. Yes. 37 1 Q. So those are qualified lawyers -- 2 A. Yes. 3 Q. -- who you assume or your evidence is they would 4 have discussed those with people who worked at 5 POL? 6 A. To the extent that HSF would have had their -- 7 they would have identified teams dealing with 8 each Rule 9 Request, with each area and they 9 would then have liaised with the relevant person 10 in the Post Office team. It wasn't dealt on 11 a basis where the partners in HSF would deal 12 with me, I would then allocate it down to 13 somebody within Post Office who then reported 14 back to me and I reported back to HSF. It 15 wasn't that relationship. 16 The relationship between myself and HSF was 17 much more at the higher level as to what we were 18 doing, what our strategic approaches were, what 19 our priorities were going to be for upcoming 20 phases. 21 Q. Is it possible that those conversations with the 22 people involved on the operational side of the 23 Post Office simply weren't taking place and that 24 you weren't aware of that? 25 A. No, because I was aware of conversations then, 38 1 because I would -- there were frustrations 2 expressed by the operational side, where they 3 would say, "I've sent material to HSF and 4 they've asked for it again and I've already sent 5 this". So there were definitely ongoing 6 discussions and ongoing exchanges, and that was 7 taking place. But the exact nature of who was 8 speaking to who and what they were asking for 9 and what was being produced was not reported up 10 to me. I was aware of the frustrations both 11 from HSF saying, "We've asked for certain 12 material which hasn't arrived". I would then 13 say, "Look what are we doing?" and then from our 14 side saying, "I've sent this three times" and 15 then me going to HSF saying, "Apparently this 16 has already been three occasions". 17 So there were conversations at the lawyer 18 level, either between the lawyers and the 19 internal lawyers, or the lawyers and HSF and the 20 operational team within Post Office. 21 Q. In order to assist those who are currently 22 working for the Post Office on their disclosure, 23 can you tell us why you think that kind of 24 a breakdown would happen? 25 A. I think it's just -- partly, I think it is the 39 1 size of the team within Post Office. There was 2 a limited number of people and some of the 3 operational staff had other BAU duties and 4 responsibilities, so they weren't dedicated 5 solely to the Post Office. And I also think 6 partly -- I'm criticising our own profession 7 here -- but I don't think lawyers sometimes 8 understand and can explain to non-lawyers what 9 they are asking for as clearly as maybe they 10 should. 11 But I also do think part of the problem can 12 be something the size of this Inquiry where one 13 member of the Post Office is responding to one 14 person in HSF, and there are three different 15 teams in HSF doing three different jobs, but the 16 Venn diagram arrangement of inquiries is certain 17 matters do overlap, and getting repeated 18 requests for a document you sent to person A 19 from person B or person C, the Post Office staff 20 get frustrated because "I've already sent that", 21 but it was to a different team for a different 22 purpose. 23 Q. So was there a lack of communication between the 24 various Herbert Smith teams rather than a lack 25 of communication between Herbert Smith and the 40 1 Post Office? 2 A. I think there was -- there may well have been 3 a lack of clear understanding but, again, the 4 sheer volume of what's being dealt with -- 5 understanding what document has come in and 6 having a central repository of all of that 7 information, plus how you name it. I name it 8 "Board meeting, 12 January", somebody names it 9 "Board meeting re Horizon", a third person names 10 it "Board meeting" -- they're asking for the 11 same document but they've called it a different 12 name because of the distinction within the 13 Rule 9 Request as to what it is you're actually 14 looking for. 15 So there may well be a lack of clarity in 16 what is being requested, consistency in namings 17 or even just an understanding within the teams, 18 Herbert Smith asking for one thing and Post 19 Office thinking they're asking for something 20 else, and I think that it is quite likely that 21 there were instances where that did happen. 22 Q. When you were in post, is it your view that Post 23 Office was lacking a central individual to 24 filter those kinds of requests or was it 25 a general understaffing or something else? 41 1 A. I think the team was generally understaffed, in 2 that there was a huge job to be done but there 3 weren't enough people to do all of the tasks, 4 which meant that you had to prioritise within 5 the teams, and that may well have led to some 6 requests or some understanding falling between 7 the cracks. 8 Q. Looking at that racial characteristics document, 9 we've dealt with it with Mr Foat, so I'm not 10 going to deal with it in depth with you today, 11 but you've said that you sat on a steering 12 committee that related to Freedom of Information 13 Act requests and that there was liaison in fact 14 between the Freedom of Information people and 15 the Inquiry people, and that I think you were 16 that figure that sat on both committees, for 17 example. 18 A. Yes. 19 Q. How, in your view, do you think it is that that 20 document, having been disclosed under the 21 Freedom of Information Act, wasn't brought to 22 the relevant people's attention promptly? 23 A. I don't -- I have no comprehension as to how 24 that did not. That would have been after my 25 time, so I don't know how that happened. I do 42 1 recall instances when documents were brought up, 2 that were going to go on Freedom of Information 3 requests and we immediately ensured HSF advised 4 the Inquiry that material was going to come out 5 because we didn't want the Inquiry's first -- it 6 may not have been relevant to an existing Rule 9 7 but it was clearly relevant to facts in issue in 8 the Inquiry and we did not want the first 9 knowledge to be the front page of the Daily 10 Mail. 11 We wanted to make sure, even if it was only 12 a day or two's knowledge, that you would have 13 known. So we did, on a number of occasions, 14 identify material that had been requested under 15 Freedom of Information requests and we bought 16 them to the attention of the Inquiry. So I know 17 that that did happen. I do not know how that 18 did not happen on this occasion. 19 Q. Did you, during your period, recognise any 20 difference in, for example, the quality of the 21 product from Freedom of Information searches, 22 vis à vis the quality that was produced for 23 Inquiry's disclosure? 24 A. No. I think my feeling was that the Freedom of 25 Information was much more directed, because they 43 1 were -- the requests would have been much more 2 specific, where the Inquiry's scope is much 3 wider and the request would be a much broader 4 brush. So what we would be looking for would be 5 a lot wider than what a Freedom of Information 6 request would be looking for. So it may well 7 have been that a Freedom of Information request 8 would divulge 14/20 documents, we would be 9 locating 15,000/20,000 documents. 10 So I do not -- as I say, I do not understand 11 how that one was located by Freedom of 12 Information when it hadn't been located 13 previously, nor how we -- the Inquiry Team were 14 not able to notify the Inquiry in advance that 15 this was coming. 16 Q. In terms of your ultimate departure, can you 17 tell us why you left the Post Office? 18 A. I think from my understanding is it was quite 19 simply budgetary, that the cost of seconding 20 a partner from a law firm in was extensive and, 21 therefore, they had to try to manage the 22 long-term budget of how to manage the fees and 23 bringing somebody in to manage the team long 24 term. And I think possibly then again the 25 initial view of you were brought in for 44 1 a short-term to deal with a certain aspect and 2 it has just blown up out of all proportion from 3 what we understood. 4 Q. We will shortly hear from Mrs Wills, who has 5 described some successes she has had in terms of 6 increasing the budget. Why do you think it is 7 or do you think it is that Mrs Wills has had 8 more success in increasing the budget than you 9 did during your time in post? 10 A. I think reality has come to play but, at the end 11 of my time, there was a much longer and much 12 more detailed understanding that small increases 13 of a bit here and a bit there were not going to 14 do it, that the Inquiry was -- and the Inquiry's 15 timetable itself had been much clearly set out, 16 that we knew then that we were looking at this 17 period, this period, this period, and hopefully 18 that the wearing down I had done during my time 19 and the work that I had done had softened the 20 way for a much clearer understanding, that the 21 budgets that had been discussed were simply not 22 valid. 23 And we had presented scenarios of increasing 24 the internal team, changing terms, changing the 25 approach. A lot of that had been done setting 45 1 out various avenues and I would hope possibly 2 then that there was a -- I've no idea, I'm 3 speculating here, which is probably incorrect 4 for me to do so -- that hopefully more funding 5 was provided from Government resources to ensure 6 that the Post Office could comply with its 7 duties and responsibilities to the Inquiry. 8 MR BLAKE: Thank you very much. 9 Sir, do you have any questions? 10 Questioned by SIR WYN WILLIAMS 11 SIR WYN WILLIAMS: Just one, if I may. Well, it may 12 be more than one but it's one discrete issue. 13 Mr Canavan, you spoke about the role of 14 counsel in the various teams which was set up. 15 I just wanted to understand what you meant by 16 that. Were you meaning that these people were 17 members of the independent bar in private 18 practice instructed by a solicitor or were you 19 talking about in-house employees? 20 A. I was discussing in-house lawyers, legally 21 qualified staff internally, not external members 22 of the bar. 23 SIR WYN WILLIAMS: Right. Okay. Well, it was only 24 one question. Thank you. Yes, thank you very 25 much. 46 1 MR BLAKE: Thank you very much, sir. We're going to 2 hear from Mrs Wills next but perhaps we could 3 take a 15-minute break, so to 11.15. 4 SIR WYN WILLIAMS: Yes, by all means. Fine. 5 MR BLAKE: Thank you very much. 6 (11.00 am) 7 (A short break) 8 (11.15 am) 9 MR BEER: Good morning, sir, can you see and hear 10 me? 11 SIR WYN WILLIAMS: Yes, I can, thank you very much. 12 MR BEER: May I call Mrs Diane Wills, please. 13 SIR WYN WILLIAMS: Yes. 14 DIANE WILLS (sworn) 15 Questioned by MR BEER 16 MR BEER: Thank you, Mrs Wills. As you know, my 17 name is Jason Beer and I ask questions on behalf 18 of the Inquiry. Can you give us your full name 19 please? 20 A. Diane Wills. 21 Q. Thank you for coming to assist us in our work, 22 and for previously providing a witness statement 23 to the Inquiry. You should have in front of you 24 a copy of that witness statement, in your name 25 and dated 22 August this year. If you turn to 47 1 the last page of it, which is I think page 37, 2 is that your signature? 3 A. It is. 4 Q. Before I ask you whether it is true, can you go 5 back to page 29, please, and look at 6 paragraph 92 at the foot of the page. I think 7 you wish to make two corrections to 8 paragraph 92; is that right? 9 A. Yes, please. 10 Q. Is the first of them by crossing out in the 11 first line the words "was agreed with" -- 12 A. Yes. 13 Q. -- and inserting instead the words "guide has 14 been designed by POL, having consulted the", 15 I'll read that again slowly: 16 "... guide has been designed by POL having 17 consulted the ..." 18 A. Correct. 19 Q. Then secondly inserting the word "internally" at 20 the end of that sentence after the word "used"? 21 A. Yes. 22 Q. So the sentence would read: 23 "A revised Rule 9 Request process guide has 24 been designed by POL having consulted the 25 external advisers and is now being used 48 1 internally." 2 A. Yes. 3 Q. Thank you. 4 With those corrections, are the contents of 5 that witness statement true to the best of your 6 knowledge and belief? 7 A. Yes, they are. 8 Q. I am going to ask you questions about a range of 9 matters referred to in your witness statement 10 but not all of them and that's because the Chair 11 of the Inquiry has read your witness statement 12 and it will be uploaded to the Inquiry's website 13 today so that everyone can see what it session. 14 I am going to focus on two principal issues, 15 firstly to seek to understand more about how 16 three disclosure failings occurred concerning 17 search terms, reviewing families of documents, 18 and the de-duplication of documents. So that's 19 a look backwards? 20 A. Mm-hm. 21 Q. Then secondly to seek to understand more about 22 what the Post Office and its advisers have done 23 to put things right. 24 A. Okay. 25 Q. Can start though with your career qualifications 49 1 and experience. I think you're a solicitor 2 having qualified in October 1997; is that right? 3 A. That's correct. 4 Q. You previously worked in the Government Legal 5 Department until in April 2012 (sic) you joined 6 the Post Office as its Legal Services Director 7 in the Remediation Unit? 8 A. April 2022. 9 Q. Did I say "2012"? 10 A. You did. 11 Q. April 2022. 12 A. Correct. 13 Q. You appear to have had two functions from April 14 2022 onwards, firstly to seek to deliver timely 15 and fair compensation to subpostmasters affected 16 by the failings of the Horizon System and 17 secondly to support the Post Office's response 18 to criminal appeals arising from the operation 19 of the Horizon System; is that right? 20 A. Correct, yes. 21 Q. With effect from beginning of January 2023 and 22 upon a promotion, you became the Post Office's 23 Public Inquiry Director; is that right? 24 A. Yes, in addition to the other role. 25 Q. You tell us in your statement that you spend 50 1 about 75 per cent of your time undertaking that 2 role and 25 per cent of the time undertaking the 3 existing role; is that right? 4 A. Yes, it's not a hard split. It'll depend on the 5 nature of the issues that are arising at the 6 time but that's roughly right. 7 Q. When you became the Public Inquiry Director, 8 that was a role previously occupied by Fintan 9 Canavan, from whom we've just heard; is that 10 right? 11 A. Correct. 12 Q. You tell us in your witness statement, no need 13 to turn it up, at paragraph 31, that you are the 14 accountable director within the Post Office for 15 its Inquiry programme? 16 A. Yes. 17 Q. Can you tell us what that means in two respects? 18 Firstly, accountable to whom? Accountable to 19 the Inquiry or accountable within the Post 20 Office? 21 A. Accountable within the Post Office. 22 Q. What does that mean: you are the accountable 23 director within and to the Post Office? 24 A. So I'm the person that would -- is responsible 25 for making sure that the Group Executive, that 51 1 the board, the Inquiry Steering Committee are 2 aware of all of the issues that are going on, 3 that I'm driving the programme forward in the 4 way I believe is needed to deliver what's 5 required for the Inquiry and so I'm, I suppose 6 internally, the face of the Inquiry, if you 7 like, for Post Office, in terms of people know 8 that I'll be there to lead the work. 9 Q. But what does accountability involve? 10 A. It's my job to deliver and, if I was not 11 delivering, then presumably that would be looked 12 at in terms of performance. 13 Q. What does the Inquiry programme mean? 14 A. So it's all the work that we're delivering 15 connected with the Post Office Horizon Inquiry. 16 Q. If we turn up your witness statement, please, 17 it's WITN09940200, and go to page 26, please. 18 At paragraph 83, you say: 19 "I inherited a very highly motivated team 20 who had been under-resourced for a long period 21 of time." 22 Yes? By "for a long period of time", do you 23 mean for the life of the Statutory Inquiry, 24 ie since at least June 2021? 25 A. So I had in mind my knowledge since I joined 52 1 Post Office, which was from April 2022 and not 2 being directly involved with the Inquiry Team 3 but just based on general knowledge and 4 understanding of the pressures that the team 5 were facing. 6 Q. So you're referring to "it had been 7 under-resourced for a long period of time", 8 meaning since at least April 2022? 9 A. From my perspective, yes. 10 Q. Because that's when you had knowledge? 11 A. Yes. 12 Q. So it had been under-resourced for at least 13 eight months by the time you arrived? 14 A. In my view, yes. 15 Q. To your knowledge, had issues as to the 16 under-resourcing of the Inquiry Team previously 17 been brought to the attention of others within 18 the Post Office? 19 A. So I don't know the specific details of what had 20 been brought to the attention of others. 21 I know, just from general conversations with 22 Mr Canavan, that there had been concerns about 23 whether the budget was at the right level but 24 I don't know more detail than that. 25 Q. To your knowledge, had issues as to the 53 1 under-resourcing of the Inquiry Team been 2 brought to the attention of the Group 3 Executive -- 4 A. I don't know. 5 Q. -- and to the board? 6 A. I don't know. 7 Q. What was your understanding of the response from 8 within the Post Office as to issues as to 9 under-resourcing having been raised? 10 A. So, as I said, I don't know that they were 11 focused specifically on under-resourcing. My 12 sense was that Mr Canavan had found it difficult 13 to enable the Post Office fully to understand 14 the extent of the obligations required, in terms 15 of a response to the Inquiry and what that would 16 entail, and that had therefore been the 17 difficulty in securing the right level of 18 budget. 19 Q. So he hadn't understood -- 20 A. Sorry, not he. He had found it hard to get the 21 rest of the Post Office to understand the nature 22 of the challenge, sorry. 23 Q. Is that the explanation for why the 24 under-resourcing had been allowed to continue? 25 A. So far as I know, yes. 54 1 Q. When was the under-resourcing of the team first 2 bought to the attention of the Group Executive 3 by you? 4 A. So I think it would have been in either February 5 or March, ahead of taking a paper to the board 6 in March, at which we focused on additional -- 7 sorry an increase in the budget and as part of 8 that additional resourcing and there would have 9 been discussions at the Group Executive level 10 prior to that. There would have also been 11 discussions prior to that at the Inquiry 12 Steering Committee, which itself has number of 13 members of the Group Executive. 14 Q. Can we look, please, at paragraph 58(d) on 15 page 18 of your witness statement? 16 SIR WYN WILLIAMS: Before you do that, unless 17 I missed it, you said February or March but 18 didn't give a year. I'm assuming it's 2023. 19 A. 2023, yes, sorry, yes. 20 SIR WYN WILLIAMS: Thanks. 21 MR BEER: You're here dealing with part of the 22 response to the Inquiry's question to explain 23 the events that led to the three disclosure 24 failings that had been identified. At (d), at 25 the foot of the page, you're dealing with the 55 1 issues of searches that have been undertaken, 2 and you say: 3 "My understanding from discussions with 4 colleagues is that Fintan Canavan and some 5 others with relevant responsibilities had 6 an ongoing dialogue with POL's external advisers 7 to seek to ensure that the searches for 8 documents and document review strategies were 9 reasonable in all the circumstances and 10 comprehensive, thorough and rigorous." 11 Then this: 12 "POL wanted to achieve cost and process 13 efficiencies if they were achievable alongside 14 fully meeting its obligations to the Inquiry and 15 its ambition to drive improvement going 16 forward." 17 So you're referring here to conversations 18 that you had with Mr Canavan about search terms 19 and review of document strategies, yes? 20 A. At a very high level. 21 Q. That's what I wanted to ask you about. This is 22 expressed at a very high level. You say on the 23 previous page the Post Office -- at the foot of 24 the page: 25 "POL [the Post Office] wanted to achieve 56 1 cost and process efficiencies if they were 2 achievable alongside fully meeting its 3 obligations to the Inquiry ..." 4 Can you explain exactly what that means, "we 5 want to fully meet our obligations to the 6 Inquiry and achieve cost and process 7 efficiencies"? 8 A. So there has never been any doubt as to Post 9 Office's support, full support for delivering 10 what's needed for the Inquiry and that 11 underpinned everything. So the board has been 12 very clear throughout, in terms of wanting to 13 ensure that everything possible was done to meet 14 the disclosure obligations. But sometimes there 15 are ways of doing things which can achieve the 16 same outcome but in a more cost efficient way 17 and that's what that's focused on. 18 Q. Can you give some examples? 19 A. So probably the biggest example I can give is 20 the -- that cost was part of the driver to 21 change the decision -- sorry, part of the 22 decision to move the recognised legal 23 representative role from Herbert Smith Freehills 24 to Burges Salmon and Fieldfisher. 25 Q. This is talking about a stage before then. 57 1 A. Yes. 2 Q. This long before then. 3 A. Yes. 4 Q. Can you give any examples about the Post Office 5 wanting to achieve cost efficiencies concerning 6 disclosure? 7 A. Only at a high level, in terms of examining, for 8 example, things like how many associates would 9 attend each of the Inquiry hearings. 10 Q. I'm talking about disclosure? This is in the 11 context -- 12 A. Disclosure, sorry. Yes, I don't have any 13 specific examples. 14 Q. What were you referring to then of POL wanting 15 to achieve cost and process efficiencies in the 16 context of disclosure? 17 A. That was how it had been explained to me in 18 terms of looking hard at how disclosure is 19 approached and seeing whether, through working 20 with Relativity, working with KPMG, there is 21 a different way that we can achieve the same 22 outcomes but at a less cost. 23 Q. Again, other than expressing it in that very 24 high level way, can you think of any practical 25 examples or actual events where that was carried 58 1 into effect or was described to you? 2 A. Perhaps the process followed -- that 3 I understood -- was followed in relation to some 4 of the hard copy work, where it was done in 5 stages, so that, for example, in the review of 6 the back-up tapes, which were discovered at 7 a site, rather than going straight into a full 8 process review, instead, the work was started to 9 identify the policies which led to them being 10 designated as back-up tapes. There was work 11 done to identify indices, so there was a high 12 level understanding of what might be on there. 13 There was then a sampling done, so it was 14 done in stages to try to achieve some 15 efficiencies whilst still being mindful of our 16 overall obligations. 17 Q. In relation to the three issues that we're 18 concerned with, the use of search terms or the 19 misuse of search terms, reviews of family 20 documents and the de-duplication exercise, was 21 the Post Office's desire to achieve cost and 22 process efficiencies relevant to any of those? 23 A. I've never heard it described in terms of a cost 24 efficiency. It was always about what was 25 reasonable to do in the circumstances, which 59 1 inevitably has some implications for cost. 2 Q. By that answer, Mrs Wills, do I understand you 3 to mean that desires to save costs are relevant 4 to the three disclosure failings that we're 5 looking at? 6 A. No, I mean that it's an inevitable output, in 7 terms of decisions taken as to what is 8 reasonable, and the approach taken, depending on 9 how wide or not that approach is taken, will 10 have different cost implications. 11 Q. Okay, I'll move on. That can come down from the 12 screen, please. One of the principal issues 13 which the Inquiry -- I think you'll 14 understand -- we're investigating, is, in the 15 substance of our hearings, the non-disclosure of 16 documents by the Post Office in civil litigation 17 and in criminal litigation, which the Post 18 Office bought against subpostmasters? 19 A. Mm-hm. 20 Q. I think you will realise, is this right, that 21 the Post Office was significantly criticised by 22 both the High Court and the Court of Appeal for 23 what was described as obstructive conduct in 24 relation to disclosure? 25 A. Yes. 60 1 Q. Against that background, we're investigating 2 it -- ie non-disclosure -- and it's been the 3 subject of significant criticism in the past. 4 Would you agree that the disclosure which the 5 Post Office gives to this Inquiry must, in the 6 circumstances, be full, accurate and timely? 7 A. Yes. 8 Q. That's essentially a commitment that various 9 senior executives in the Post Office have been 10 committed to giving in the Inquiry; is that 11 right? 12 A. Yes. 13 Q. And it's what the law requires? 14 A. Yes. 15 Q. Can we look, please, at paragraph 43 of your 16 witness statement on page 13. Just explaining 17 the abbreviations, HSF is a reference to Herbert 18 Smith Freehills -- 19 A. Yes. 20 Q. -- the Inquiry lawyers and, until recently, the 21 Inquiry recognised legal representatives -- 22 A. Yes. 23 Q. -- for the Post Office. P&P are Peters & 24 Peters. We will come to each of those in 25 a moment. You say: 61 1 "[Herbert Smith Freehills] and [Peters & 2 Peters] had a broad, general authority from [the 3 Post Office] as its instructed representatives 4 to progress disclosure, in conjunction KPMG 5 International, without reverting to [the Post 6 Office] for specific instructions." 7 A. Yes. 8 Q. You're here referring to, I think, the past, 9 ie the period before you took up office on the 10 1 January 2023; is that right? 11 A. That's the system I inherited on joining, and 12 which, to an extent, is continued until more 13 recently. 14 Q. To your knowledge, was the decision taken by the 15 Post Office to give Herbert Smith Freehills and 16 Peters & Peters a broad, general authority, 17 without reverting to the Post Office for 18 specific instructions in relation to disclosure 19 issues, a deliberate decision? 20 A. I don't think I can answer that. I wasn't 21 involved in those discussions. 22 Q. The consequence of the decision, would you 23 agree, that if there are issues or problems with 24 the way that disclosure is undertaken by those 25 external advisers, then the Post Office can say, 62 1 "Although ultimate responsibility rests with us, 2 Post Office, to give proper disclosure, in fact 3 the errors and mistakes were not made by us they 4 were made by others"? 5 A. I don't think that was ever the intention behind 6 it. 7 Q. That's the consequence of it? 8 A. Yes. 9 Q. Indeed, I think that's exactly what you do in 10 your witness statement, isn't it? 11 A. Yes. 12 Q. If we just look at paragraph 10 on page 4. 13 You're dealing with Request 1 at paragraph 10, 14 which was: 15 "Please explain the events which led to the 16 three disclosure issues, explaining in detail 17 all relevant decisions and communications that 18 led to the errors and identifying those 19 involved." 20 You say: 21 "The relevant decisions and communications 22 took place between POL's external advisers. 23 [The Post Office] did not direct these 24 communications or take these decisions. 25 Consequently ... the witnesses from [the Post 63 1 Office's] external advisers will be able to 2 provide closer accounts." 3 Then over the page at page 12, you're 4 dealing with Request 2.b. Request 2.b was: 5 "Please set out where you consider 6 responsibility for the errors lie." 7 You say at 12: 8 "As I have explained in relation to 9 Request 1, the relevant decisions and 10 communications took place between [the Post 11 Office's] external advisers. Accordingly, any 12 responsibility primarily sits with [the Post 13 Office's] external advisers." 14 Yes? 15 A. Yes. 16 Q. In relation to other issues, signing off 17 proposed communications to the Inquiry, signing 18 off written, closing and oral submissions to the 19 Inquiry, signing off submissions in relation to 20 compensation issues, you indicate in your 21 witness statement that by contrast, you were 22 involved very directly with those? 23 A. Yes, either I or my team. 24 Q. You tell us that, in relation to those issues, 25 communications with the Inquiry, signing off 64 1 submissions and signing off communications in 2 relation to compensation issues, you have 3 frequent and direct contact with Herbert Smith 4 Freehills lawyers? 5 A. Yes. 6 Q. You receive updates from them via email? 7 A. And through oral conversations. 8 Q. You say that -- in your witness statement -- you 9 attend a series of standing meetings with them? 10 A. Yes. 11 Q. Why, to your knowledge, was a different, more 12 hands-off approach taken in relation to 13 disclosure issues? 14 A. So reference to the communications with the 15 Inquiry, I had in mind the production letters 16 that go to the Inquiry so that Post Office has 17 visibility at that stage of what is being 18 produced to the Inquiry. We have the earlier 19 visibility of the Rule 9s coming in. Those 20 discussions -- 21 Q. Sorry to interrupt. 22 A. Sorry. 23 Q. What about the bit in the middle? 24 A. That's what I was just going to come to. The 25 regular calls that we have with Herbert Smith go 65 1 through the Rule 9s that are currently 2 outstanding. There is a high level discussion 3 so that I'm aware of any resourcing issues, any 4 major issues. What I didn't have, and I think 5 my team only had a limited knowledge of, was 6 precise details, such as things like search 7 terms and how de-duplication was being applied. 8 Q. My question was: you get sight of the Rule 9 9 that comes in, you get sight of the production 10 at the end that goes back to the Inquiry, why 11 was a more hands-off approach taken in relation 12 to the part in the middle? 13 A. So I think it's a combination of two reasons. 14 One is that the reason we appointed the firms we 15 did was because of their expertise and we wanted 16 to be able to make good use of that and also 17 their greater resource levels than we had, but 18 also that there wasn't sufficient capacity 19 within the team to have the level of engagement 20 on these issues that I would have liked. 21 Q. So it comes back to the resourcing of the 22 internal POL Inquiry function? 23 A. In part, yes. 24 Q. You speak in your witness statement about 25 a series of committees and groups within the 66 1 Post Office that have responsibility for the 2 Inquiry programme. If I can deal with each of 3 those in turn, there's something called the 4 Inquiry Team -- capital "I", capital "T" -- at 5 the Post Office made up of lawyers and project 6 managers. Is it right that that had no 7 responsibility for or oversight of any 8 disclosure issues? 9 A. No, that's not correct. 10 Q. There was and is an Inquiry Steering Committee? 11 A. Yes. 12 Q. Is it right that that had no responsibility for 13 or oversight of disclosure issues? 14 A. No, it did have. 15 Q. Sorry? 16 A. It did have. 17 Q. It did have? 18 A. Yes. 19 Q. Can you tell us, in relation to each of those, 20 the levels of firstly the Inquiry Team, the 21 oversight of and responsibility for disclosure 22 issues? 23 A. So within the team, each of the lawyers had 24 a particular responsibility for a phase of the 25 Inquiry and any of the -- any Rule 9 issues that 67 1 arose in relation to that phase, they would have 2 the day-to-day responsibility for contact with 3 the external lawyers. They were supported in 4 that role by the project management team and 5 a specific person in the team who had 6 responsibility due to their longstanding 7 corporate memory of helping us understand who, 8 across the business, we may want to engage with, 9 in order to gain information -- corporate 10 knowledge from subject matter experts. 11 Q. If your paragraph 43 is right, that Herbert 12 Smiths and Peters & Peters had a broad, general 13 authority in relation to disclosure, without 14 having to revert to the Post Office for specific 15 instructions, what oversight and responsibility 16 was it necessary for these Inquiry lawyers 17 within the Post Office's Inquiry Team to 18 undertake? What were they doing about 19 disclosure? 20 A. So my understanding is that the team's main role 21 was in order to help HSF understand who may have 22 the relevant data that was relevant, where that 23 might be held, to try to find relevant 24 custodians within the department. We didn't, 25 for example, have any knowledge about 68 1 an approach being taken on family documents. So 2 that wasn't something that we would have had 3 knowledge or oversight of nor, as I say, the 4 de-duplication, other than, I guess, at a high 5 level, that it makes sense to try and avoid 6 providing duplicates of something, but no more 7 than that. But the team's role was much 8 involved in just helping source the material. 9 Q. Where are the documents; what are our 10 repositories; which buildings need to be looked 11 at; that kind of thing? 12 A. Yes, but then also through discussions with 13 those individuals to try to help build the 14 information necessary to -- for HSF to then 15 determine search terms. So, for example, what 16 project names were used, you know, what common 17 acronyms or abbreviations were used within 18 a team, so that that would then help HSF to make 19 the more technical decision about precisely what 20 search terms would be used. 21 Q. I see. So narrowing it down, then, the Inquiry 22 Team didn't have any involvement in or knowledge 23 of the detail of decisions taken as to the 24 review of family documents, which is one of the 25 areas we're looking at -- 69 1 A. That's correct. 2 Q. -- and didn't have any knowledge of or 3 involvement in decisions concerning the 4 de-duplication of datasets, other than to know 5 in general terms it's a good idea that things 6 are de-duplicated -- 7 A. Correct. 8 Q. -- the second error we're looking at. 9 On the third issue, search terms, I think 10 you just told us that they did -- the Inquiry 11 lawyers within the Post Office -- have 12 discussion and liaison with the external 13 advisers over the appropriate search terms to 14 use; is that right? 15 A. No, I think it's more that they acted as 16 a conduit for HSF to engage directly with the 17 business and may have been present on calls, for 18 example, but were not themselves, so far as I'm 19 aware, directing the nature of the precise 20 search terms. 21 Q. Can we ask the same set of questions in relation 22 to the Steering Committee, then? Any 23 involvement on or responsibility for the family 24 review issue? 25 A. They'd have had no knowledge of that. 70 1 Q. In relation to the de-duplication issue? 2 A. Again, no knowledge. 3 Q. And search terms? 4 A. I wouldn't expect them to have any knowledge of 5 that either. The knowledge for SteerCo about 6 the Rule 9s was a broad sense of what was coming 7 in, what was being looked at, what we were 8 finding, any challenges in that, timescales, 9 whether it had impact on resources. It was more 10 strategic involvement than day to day. 11 Q. Same answers in relation to the Group Executive? 12 A. Yes. 13 Q. Still further up the tree, same answers in 14 relation to the Post Office Board itself? 15 A. Yes, that's correct. 16 Q. Would you say that through the Inquiry team 17 within the Post Office, it -- the Post Office -- 18 exercised intrusive supervision and intrusive 19 oversight of the disclosure process that was 20 being carried out in its name? 21 A. I don't think I could categorise it as that, no. 22 Q. Can we turn to paragraph 13 of your witness 23 statement, please, which is on page 5. We've 24 dealt with paragraph 12, responsibility 25 primarily resting with external advisers. You 71 1 say in 13: 2 "However, that doesn't change the fact that 3 [the Post Office] remains responsible for its 4 own disclosure to the Inquiry. [You] hope it is 5 made clear from the rest of this statement that 6 [the Post Office] takes that responsibility 7 extremely seriously." 8 So you rightly say that the Post Office 9 remains responsible for its disclosure to the 10 Inquiry and you say that it takes the 11 responsibility "extremely seriously". Why, 12 given those two things, did the Post Office not 13 exercise intrusive supervision or oversight of 14 the disclosure that we're being given or not 15 being given? 16 A. I think until recently, until the issues have 17 come to light, it wasn't understood that there 18 were issues of the scale that we've now come to 19 realise and so I don't think there was -- it was 20 understood that there was a concern about the 21 approach being taken, that meant more intrusive 22 supervision was required. 23 Q. But Mrs Wills, in a case concerning significant 24 non-disclosure to criminal courts in the past, 25 which has been extensively and deeply criticised 72 1 by both the High Court and the Court of Appeal, 2 wouldn't the disclosure being given to this 3 Inquiry naturally require intrusive supervision 4 and oversight, whether or not you knew there was 5 a problem? 6 A. Yes, I can see that but I believe Post Office 7 considered it was meeting its obligations by 8 having appointed the firms it had. 9 Q. One of the issues we're investigating is 10 delegation of functions to a supposedly expert 11 contractor, Fujitsu -- 12 A. Mm-hm. 13 Q. -- and whether washing one's hands of a problem 14 and saying, "It's down to our external 15 contractor", was an appropriate way to go. Are 16 there shades of that on this occasion in 17 relation to disclosure? 18 A. It's not how I'd see it. The team had regular 19 engagement with Herbert Smiths, it had regular 20 engagement with KPMG and Peters & Peters and 21 I think the belief was we had sufficient 22 knowledge of how disclosure was being conducted. 23 Q. Can we turn to page 10 of your witness 24 statement, please. You say in paragraph 34: 25 "My focus has been in 3 main areas ..." 73 1 Then at (c), you say: 2 "Ensuring the ISC [that's the Inquiry 3 Steering Committee] and [the Post Office's] 4 Group Executive and Board are appropriately 5 aware of and consulted on risk issues relating 6 to the Inquiry as they are identified. This is 7 obviously essential to thorough risk governance 8 ..." 9 In order for thorough risk governance to 10 take place, don't you have to know what the 11 risks are in the first place? 12 A. Yes. 13 Q. Does that not necessarily involve conversations 14 and written communications with your advisers 15 that seek to identify what the risks are? 16 A. I believe Post Office thought it understood the 17 nature of the risks and it's only really since 18 June, with all the extensive work that's been 19 undertaken, that we've identified additional 20 matters. 21 Q. So if it's the case that the external advisers 22 have adopted a disclosure strategy that's mainly 23 focused on keyword searches, the conversation 24 is, "We've adopted a disclosure strategy that's 25 mainly focused on keyword searches, these are 74 1 the risks that that involves" or "We've decided 2 to leave it to the discretion of an individual 3 low-level reviewer to decide on a case-by-case 4 basis whether to look at a family of documents, 5 that carries the following risks"? 6 Those things need to be identified, don't 7 they, they need to be bubbled up to the surface 8 in order that risk governance can take place, 9 don't they? 10 A. Yes, they do. 11 Q. By the time of your arrival, that hadn't 12 happened, had it? 13 A. I can't speak about what had happened before my 14 arrival. 15 Q. Can we turn to paragraph 40 on page 12, please. 16 In the middle of the paragraph here, and you're 17 talking about the current day, you say: 18 "I attend twice weekly standing meetings 19 with the 3 partners responsible within [Herbert 20 Smith Freehills] for leading the Inquiry work as 21 well as regular detailed strategy meetings. 22 These meetings are used to discuss priority 23 issues and risks as they are identified and to 24 establish matters that require escalation ..." 25 A. Yes. 75 1 Q. Between January and the discovery of problems 2 with disclosure, in late May and early June 3 2023, were the risks associated with the way in 4 which Herbert Smith Freehills, Peters & Peters 5 and KPMG were approaching disclosure ever 6 discussed in the meetings you refer to here? 7 A. Yes. 8 Q. What risks were identified? 9 A. So there was frequent discussions about the 10 ongoing work in relation to hard copy data and 11 some of the repositories that had been found 12 there, including electronic data. I was also 13 aware, around April time, of concerns that -- as 14 to whether or not the right repositories had 15 been searched in relation to one particular 16 Rule 9. I can't immediately think of other 17 specific risks. 18 Q. Again, that's focused on places and vaults of 19 material -- 20 A. Yes. 21 Q. -- rather than the three issues we are looking 22 at. Had any discussion occurred between January 23 and late May/early June 2023 about any risks 24 being associated with the way in which those 25 three advisers were approaching disclosure 76 1 concerning the three risks that have now been 2 identified? 3 A. I had been made aware of some concerns as to 4 resourcing challenges within KPMG, that could 5 have impacted on disclosure and we'd been 6 addressing those but I'd not been made aware of 7 any -- as far as I can recall -- certainly 8 nothing about family documents or 9 de-duplication. I don't recall being made aware 10 of any specific concerns about search terms. 11 Q. What were the resourcing concerns within KPMG 12 that were drawn to your attention between 13 January and end of May 2023? 14 A. I think in about February or March, concerns 15 were reported through to my operations and 16 strategy director that the time being taken by 17 KPMG to respond to certain requests was slowing 18 things down and impacting the ability of HSF to 19 carry out the reviews that were needed. So Post 20 Office worked with KPMG to develop two 21 solutions: one an offshoring resource centre in 22 India that KPMG could then call on to boost 23 their resources where required; and also 24 a system in place to ensure that it had the 25 capability -- sorry, the capacity to do weekend 77 1 working and late evenings where particular 2 Rule 9s required it. 3 Q. At that time, were KPMG not working in the 4 evenings or at weekends? 5 A. I don't know the detail, I'm afraid, but 6 I assume there was a need for more, given the 7 conversations that we had. 8 Q. Was the offshoring to India solution taken up? 9 A. Yes. 10 Q. What is the nature of the offshoring to India 11 solution? What is done? 12 A. I don't know the technical detail of the work 13 they do but I know there is an additional team 14 based out there, I think in the region of around 15 about 12 people, who supplement the work of the 16 London-based office, dealing with some of the 17 more simple tasks, as I understand it. 18 Q. Do you know what that is? 19 A. I don't, I'm afraid. 20 Q. Have you any clue as to even the realm within 21 which the offshore India team is operating? 22 A. Sorry, could you clarify your question? 23 Q. Yes, are they doing de-duplication? Are they 24 doing data reconciliation? 25 A. I'm sorry, I don't know. Hopefully Mr Tombleson 78 1 will be able to answer that this afternoon. 2 Q. Can we turn forward, please, to page 30 of your 3 witness statement and paragraph 93. You're 4 again talking in the present here. You say: 5 "The [Inquiry Steering Committee], which 6 I chair, meets at least fortnightly and 7 disclosure ([both] remediation and going 8 forward) is part of the regular agenda." 9 When you joined, is it right that disclosure 10 was not part of the regular agenda of the ISC? 11 A. It wasn't a standing agenda item, in the way 12 it's become since June but, certainly, any 13 issues that were arising, such as in relation to 14 hard copy data, were brought to the attention of 15 SteerCo, as was needed. 16 Q. Do you know why disclosure was not part of the 17 standing agenda before the disclosure failures 18 which have been revealed were revealed? 19 A. So the papers that go to SteerCo comprise 20 a number of components. There are papers which 21 are for noting on issues we just want to make 22 SteerCo aware of, there are decision papers, and 23 then there's a full suite of management 24 information, which includes, for example, 25 progress on Rule 9s and also has management 79 1 information relating to progress generally, in 2 relation to disclosure. So the papers were 3 always there; they just may not have always been 4 discussed. 5 Q. Why was that? 6 A. If there was a particular issue that we needed 7 a view on, we would have discussed it. If 8 members of SteerCo had any questions in relation 9 to the papers we were providing, we would have 10 discussed it. Otherwise, it was there for their 11 general awareness. 12 Q. Standing back, the disclosure given by the Post 13 Office to this Inquiry is one of the main 14 features of its engagement with the Inquiry. 15 It's not necessarily the advocacy that Post 16 Office does within the Inquiry. The provision 17 of witness statements and disclosure are the 18 main touchpoints between the Post Office and the 19 Inquiry. The impression that I am getting from 20 reading your witness statement -- please correct 21 me if I'm wrong -- is that there was a -- before 22 your arrival and to some extent for a period 23 after your arrival, disclosure, although it was 24 spoken about in terms of being an important and 25 high priority, a rather laissez-faire attitude 80 1 was taken to it in terms of that's the 2 responsibility of the external advisers and not 3 us; is that unfair? 4 A. I think it is unfair. I think I categorise it 5 is that part of our relationship with the 6 external advisers is we would expect them to be 7 bringing to our attention any issues which we 8 needed to address or be concerned about and then 9 we'd have the regular updates and then, between 10 those two things, we felt we were getting 11 sufficient oversight of information. 12 Q. So reliance on the external advisers to identify 13 problems with their own work? 14 A. Problems with their own work but also challenges 15 they may have been facing within Post Office, 16 for example. Getting access to right 17 repositories, et cetera. 18 Q. Putting it frankly, isn't this the case, isn't 19 this the Inquiry where the internal Post Office 20 team should be all over disclosure from the top 21 and the bottom of it, day in, day out? 22 A. Yes, in ideal world, and I think that's what I'm 23 now driving towards and it's a wholesale change 24 since June. 25 Q. I appreciate that you're, to some extent, 81 1 speaking of the past here and that places you in 2 some difficulty but shouldn't that have been 3 obvious from the start? Why did it take these 4 three disclosure failings to move disclosure up 5 the agenda? 6 A. So I think it's been a trajectory of 7 understanding on the part of the Post Office, in 8 the way Mr Canavan described, in terms of Post 9 Office not having that initial understanding of 10 what's required by an Inquiry and that gradually 11 increasing, and the scale of the issues getting 12 broader, the scale of the challenges being faced 13 getting bigger and it coming right up the 14 agenda. 15 Q. Can I turn, then, to the three issues identified 16 by the Chair in his previous decisions and 17 directions on disclosure failings by the Post 18 Office: firstly the use of search terms; 19 secondly, decisions made to review or not to 20 review families of documents; and, thirdly, 21 de-duplication exercises carried out on behalf 22 of the Post Office. 23 So, firstly, the approach taken to the use 24 of search terms and the searching strategy. Can 25 we begin by looking at paragraph 58(a) on 82 1 page 17 of your witness statement. 2 A. Yes. 3 Q. You tell us: 4 "My understanding, based on discussions with 5 colleagues and information provided by [the Post 6 Office's] external advisers, is that the POL 7 employees/contractors were not involved in 8 data-strategy discussions at a detailed level. 9 In general: 10 "(a) [The Post Office] was updated about 11 some of its external advisers' search parameters 12 and review strategy choices as part of 13 business-as-usual client reporting. My 14 understanding is that [the Post Office's] 15 external advisers' decisions about search 16 parameters and review strategies largely 17 developed iteratively in response to feedback 18 and knowledge obtained through document searches 19 and document review. [Post Office] 20 employees/contractors were not substantively 21 involved in discussions about those detailed 22 matters because [the Post Office's] external 23 advisers had the specialist experience and 24 resource and, because they were conducting the 25 disclosure exercise on a day-to-day basis, had 83 1 the detailed knowledge required to make those 2 choices, whereas [the Post Office] did not." 3 So you're essentially saying three things 4 there. Firstly, it was the external advisers 5 that made the decisions about search parameters 6 and review strategies, yes? 7 A. Yes. 8 Q. Secondly, Post Office employees were not 9 involved in the substance of such decisions 10 although they were sometimes told about them? 11 A. So, as I said earlier, they may have been 12 involved in discussions to help the external 13 advisers understand what search terms they 14 should apply in terms of, for example, a project 15 name but I don't believe they were involved in 16 very specific search terms to be used generally 17 in relation to a Rule 9. 18 Q. Thirdly, they weren't so involved at any 19 detailed level because it was believed that the 20 external advisers had the requisite knowledge 21 and the Post Office didn't? 22 A. Yes. 23 Q. Can we look at an example. It's, in fact, the 24 search terms which turned up what is now known 25 as Appendix 3. 84 1 A. Yes. 2 Q. Can we look, please, at POL00114170DS, page 82. 3 We'll see this is the first interim disclosure 4 statement dated 27 May 2022 and if we can go to 5 page 82 and look at the bottom, please. 6 So we'll see in the third column across, 7 under A, a copy of the relevant request. Here: 8 "Copy of [Post Office] Investigations Policy 9 (together with all iterations of the same since 10 1999 that are within [Post Office's] custody and 11 control)." 12 That was the Inquiry's request, yes? 13 A. Yes. 14 Q. Then we'll see the explanation of what is given 15 against A in the far right-hand column: 16 "The [Post Office] produced approximately 17 260 documents in response to the Inquiry's 18 request for all iterations of the investigations 19 policy from 1999 to date. In order to identify 20 these materials, the following search was 21 conducted over all of the materials which the 22 Post Office had at that time harvested from the 23 Postal Museum and Oasis archives ..." 24 Over the page, if you look in the right-hand 25 column: 85 1 "Search Terms: 'Policy' AND ('Investigat*' 2 or 'Prosecut*' or 'Whistle')." 3 So they were the search terms that were 4 used, yes? 5 A. Yes. 6 Q. So the search term that the adviser, the lawyer, 7 used here was the word policy, where it appeared 8 in the same document as "Investigat*" or 9 "Prosecut*" or "Whistle", yes? 10 A. That's my understanding based on that document, 11 yes. 12 Q. Yes. So what has been done is the truncated 13 words of "investigate", "investigation", 14 "investigator", or "prosecution", or 15 "prosecutor" or "Prosecuting" have been used. 16 A. Yes. 17 Q. You can see the little "*" after the T, yes? 18 A. Yes. 19 Q. So these search terms wouldn't catch documents 20 if they were called "protocols"? 21 A. If these were the only search terms used, that's 22 correct. 23 Q. They wouldn't catch documents if they were 24 called "guides" -- 25 A. No. 86 1 Q. -- or if they were called "guidance"? 2 A. No, I think that's correct. What I don't know 3 is whether other searches were being done to try 4 to identify those materials. 5 Q. Not so far as we've been told. Isn't that the 6 kind of thing that a Post Office people might be 7 able to tell the lawyer, "But hold on, I've got 8 some knowledge of these things and, in the Post 9 Office, we didn't call things policies or we did 10 call things policies but there was this other 11 level of document as well that was called 12 a guidance. You need to use the truncated term 13 guide or guidance". 14 A. So that's my understanding of the role of our 15 subject matter experts within the business, is 16 that's how we would use them. I'm not familiar 17 with any particular discussions that were 18 conducted in relation to this or other searches. 19 Q. What you told us earlier in your statement was 20 that there wasn't any detailed liaison over 21 search terms involving Post Office employees, 22 subject matter experts? 23 A. I think what I tried to explain was that the 24 role of the subject matter expert would be, as 25 you've said, to sort of say, did we use policy, 87 1 did we use guidance? What I wouldn't expect 2 them to do is advise, "That means you therefore 3 need to use this truncated word in order to find 4 all the documents". It's that balance. 5 Q. To your knowledge, did that first part of the 6 conversation ever occur? I know you won't know 7 in relation to this but, at a general level, 8 "We've got this Rule 9 in, we need to get the 9 subject matter expert into the room, let's sit 10 down with them for a day and talk through it, 11 work out which search terms they would suggest 12 are used because, after all, they've spent their 13 life dealing with investigations policy or 14 prosecution policy"? 15 A. Yes. 16 Q. Did that occur? 17 A. So I've never been directly involved in any of 18 those discussions but that's certainly the -- 19 I've always understood those to happen. What 20 I've sought to do is make that more systematic 21 so that we now have a list of identified subject 22 matter experts across the business who I believe 23 would be able to help us and, whenever any new 24 requests for information now comes in, one of 25 the very first things we do in an initial 88 1 strategy meeting is to discuss with them -- to 2 have those kind of discussions. 3 So I understand it was happening, but 4 I don't know whether it happened in every single 5 case. 6 Q. Hold on. You told us back in paragraph 58(a) of 7 your witness statement that, although the Post 8 Office was updated about some of the search 9 parameters and review strategies, the Post 10 Office was not substantively involved in 11 discussions about any detailed matters 12 concerning search parameters and review 13 strategies, because the lawyers, the expert 14 advisers, had the specialist knowledge and Post 15 Office did not. I think you've just said that, 16 in fact, on occasion, Post Office people did 17 have such specialist knowledge and they were 18 brought into the conversations? 19 A. So I'm sorry if I'm not explaining this clearly. 20 I am seeing this in two levels. One, that 21 there's a general discussion with the people 22 whose role it was to know about these sorts of 23 documents, so that conversations could be had to 24 enable the lawyers to understand what sort of 25 things they should look for. What I wouldn't 89 1 expect them to have been involved in is the 2 actual construct of the search terms on the back 3 of that, and that's -- it's that different layer 4 that I was trying to get across. 5 Q. Okay. So how frequently were they involved in 6 discussions over the appropriate search terms to 7 use, in relation to all -- 8 A. So the second layer that I'm -- 9 Q. No, the first layer? 10 A. The first layer? 11 Q. Yeah. 12 A. I couldn't quantify it because I think it would 13 depend entirely on the nature of the Rule 9, the 14 extent to which there were people within the 15 Post Office that had the right corporate 16 knowledge, the earlier in time things go back, 17 obviously the harder that was. I'd expect it to 18 be much more frequent, going forward, as the 19 date range moves forward. 20 Q. So if we wanted to, we could drill down into 21 each Rule 9, if we asked for disclosure, to see 22 the extent to which a subject matter was brought 23 in to the room, and did help the lawyers to 24 develop search terms? 25 A. Would assume those records would exist, yes. 90 1 Q. Why has it increased in frequency more recently? 2 A. Because the nature of the -- as we move forward 3 in time, there's more likely to be people within 4 Post Office that do have that corporate 5 knowledge, and so are able to bring more 6 assistance. 7 Q. Why, as we move forward in time, is there more 8 likely to be people with corporate knowledge? 9 A. Just because other people might have left, in 10 some of the very earlier stages. 11 Q. Oh, I see. So you're not saying that the longer 12 the Inquiry goes on, the more knowledge? 13 A. No, no, no. 14 Q. -- will be obtained? 15 A. No, sorry. 16 Q. Understood. Can we look please at paragraph 21 17 on page 38 of your witness statement. We should 18 look at the foot of page 30, please, to get some 19 context. Page 63 at the foot. You're here 20 dealing with Rule 9 Requests 11 and 14. These 21 are the relevant requests, certainly the second 22 of them that ought to have -- it's accepted, 23 I think -- turned up not only Appendix 3 but 24 Appendix 6. You say: 25 "I understand from [Peters & Peters] and HSF 91 1 that both firms were involved in deciding the 2 search terms in relation to ... requests 11 and 3 14, with HSF having primary responsibility in 4 respect of some of the questions and [Peters & 5 Peters] having primary responsibility in respect 6 of others. The search terms used to seek to 7 identify [the Post Office's] policies and 8 procedures relating to prosecutions and criminal 9 investigations (ie broadly those relevant to 10 questions 15 and 46 of [Rule 9(11)] and question 11 18 of Rule 9 Request 14) were significantly 12 informed by work [Peters & Peters] had done in 13 connection with the post-conviction disclosure 14 exercise (utilising their pre-existing knowledge 15 and understanding of [Post Office's] documents). 16 I understand there was also liaison between HSF 17 and [Peters & Peters] in relation to appropriate 18 search terms, particularly in relation to 19 questions 15 and 46 of Rule 9 Request 11." 20 What you don't say there is that there was 21 any involvement by anyone from the Post Office? 22 A. I'm not aware of whether there was. 23 Q. Did you ask? 24 A. (Pause) 25 I think I must have done. I don't know why 92 1 I haven't mentioned it here, I'm sorry. 2 Q. Does the absence of any positive return in this 3 paragraph suggest that there was no involvement 4 by the Post Office in the development of search 5 terms concerning these parts of Rule 9 requests 6 11 and 14? 7 A. So I know there were discussions with the 8 security team but I don't know whether that was 9 done as part of the remediation process or 10 whether it was done originally. I'm afraid 11 I can't -- I'm not entirely sure. 12 Q. You say here that the search terms used were 13 significantly informed by the work Peters & 14 Peters had done with the post-conviction 15 disclosure exercise. The search terms that were 16 used were truncating the word "investigation" to 17 the word "investigat". How was the work that 18 they had done in connection with the 19 post-conviction disclosure exercise utilised in 20 order to truncate the word "investigation" to 21 the word "investigat"? 22 A. I don't know, I'm sorry. 23 Q. You've said here that it's because of their past 24 knowledge of the post-conviction disclosure 25 exercise that they developed these search terms. 93 1 What we've seen is that, for two words, they cut 2 the end off. What's the relationship between 3 this past specialist expertise and cutting the 4 end of the word? 5 A. So I don't know if there's a relationship with 6 cutting the end of the word. I imagine it's 7 with identifying the word initially, so the word 8 "investigation", not that it would appear to 9 require much specialist knowledge. 10 Q. We asked for "investigations policy"? 11 A. I can't answer that, I'm sorry. 12 Q. What did you mean by this, that the search terms 13 used to seek to identify policies and procedures 14 relating to prosecutions and investigations were 15 informed by the work they'd already done? 16 Because we've seen the search terms and all they 17 did was cut the end off a couple of words? 18 A. Yes, so probably a more accurate description is 19 that the work, generally, that Peters & Peters 20 had done, for the post-conviction disclosure 21 exercise, had been utilised in order to find 22 requisite documents, build repositories, and 23 their knowledge of that work was then used in 24 terms of helping to develop search terms 25 generally, not necessarily in relation to this 94 1 specific aspect. 2 Q. I understand. Can we lastly, on this part of 3 the issue, look at paragraph 62 on page 20 so 4 the previous page. Thank you. You say: 5 "Search terms were selected by HSF or 6 [Peters & Peters] (or through collaboration 7 between the 2 firms) and implemented by ... 8 KPMG." 9 Then you make this slightly more positive 10 assertion: 11 "[The Post Office] was not involved in 12 substantive discussions about what search terms 13 should be used ..." 14 So that does sound like you asked the 15 question in relation to these Rule 9 requests 16 and were given the answer, that the Post Office 17 wasn't involved in discussions about what search 18 terms should be used? 19 A. Possibly. It may be back to the same point 20 I was trying to make earlier about the higher 21 level of involvement of Post Office in 22 understanding the nature of work to enable those 23 search terms to then be devised by the lawyers, 24 and I was meaning that there was not involvement 25 in the sort of end product as to precisely what 95 1 search terms should then be used. 2 Q. So the cutting the end off words? 3 A. Yes. 4 Q. You say in brackets at the end there: 5 "(... I have been informed by colleagues 6 that some lists of search terms were shared with 7 POL)." 8 What was the purpose of sharing some lists 9 of search terms with the Post Office? Why were 10 they being shared? 11 A. I assume so that there was a record for Post 12 Office so that it could see what had been done. 13 Q. Was there ever any request from the external 14 advisers, "These are the search terms that we've 15 settled on, have you any comments to make"? 16 A. I'm not aware of any. 17 Q. Here you're saying that the Post Office was not 18 involved in substantive discussions about search 19 terms. Do you accept or acknowledge that now to 20 have been a mistake? 21 A. So I think there will always be a certain level 22 of expertise that we'd expect -- or to rely on 23 our external advisers for in terms of the best 24 way to construct search terms to deliver the 25 best results from Relativity. Should we be as 96 1 involved as we can be in terms of identifying in 2 the first place what search terms to use? 3 Absolutely. 4 Q. Do you know why it didn't happen? 5 A. No. 6 Q. In paragraph 64 over the page, please, you say: 7 "I understand from [the Post Office's] 8 external advisers that Appendix 6 [that's the 9 identification codes document] was not returned 10 by any of those search terms (or, indeed, any 11 other search terms that have been implemented). 12 From information I have been provided with by 13 [the Post Office's] external advisers as part of 14 the remediation process, I understand that this 15 is because the document really is quite 16 exceptional and does not contain any of the 17 words that would normally be seen in a policy or 18 procedural guidance." 19 Is that the explanation that has been sort 20 of settled and accepted between the Post Office 21 and its external advisers, because the 22 identification codes documents doesn't contain 23 words that you would expect to see in a policy 24 or procedural guidance, we couldn't reasonably 25 expect to have found it? 97 1 A. I think it's one part of the explanation. 2 Q. What about the other part of the explanation? 3 A. So I think that then takes you to the need to 4 discuss with subject matter experts and then 5 also in relation to the approach to family 6 documents. 7 Q. So a discussion with the subject matter expert 8 might have revealed, for example, "Well, hold 9 on, we have these compliance standards that we 10 had to achieve in the security team, and they 11 were sent out regularly by email in a series of 12 attachments in a zip file. Some of them 13 concerned file completion, some of them 14 concerned the need to insert an identification 15 code. You need to look for the zip file which 16 has got that suite of documents in it". 17 That's the kind of thing that might have 18 been discovered if you bring people whose day 19 job it was for decades to conduct investigations 20 and prosecutions into the room, isn't it? 21 A. Yes. The reason I'm hesitating is because that 22 document was on Relativity, so it had been 23 harvested. I don't know the circumstances which 24 had led to that document being harvested, so 25 I don't know if it was the failure to discuss 98 1 with subject matter experts which led to it 2 particularly being identified. 3 Q. Well, to that extent, this explanation here 4 misses the point, doesn't it, because, as you've 5 rightly said, Appendix 6 was part of a suite of 6 policy and procedure documents. 7 A. Yes. 8 Q. One part of the suite was responsive to a search 9 term but nobody looked, it seems, at the other 10 documents in the suite of which the responsive 11 document was a part? 12 A. Yes, but I was tying to answer the question in 13 relation to search terms. 14 Q. But, overall, would you accept that the real 15 lesson to be taken from this episode concerns 16 the need to look at other documents within 17 a suite of or a family of documents of which 18 a responsive document is a part? 19 A. Yes, and I understand that, you know, there was 20 no hard and fast rule that families would never 21 be looked at. It was always decisions were 22 taken as to the circumstances in which it would 23 be appropriate. I'd expect that to be looked at 24 hard, and even harder now, as to the right 25 approach. 99 1 Q. So you would expect what to have been looked at 2 hard? 3 A. In relation to whether or not the full suite of 4 documents needed to be reviewed. 5 Q. We're going to come to that in just a moment but 6 just finishing off search terms, if we can 7 lastly look at paragraph 114 of your witness 8 statement, which is on page 36. You're here 9 summarising your position on where 10 responsibility lies for the non-disclosure by 11 reason of search terms. At paragraph 114, you 12 say: 13 "I do not have the knowledge or technical 14 expertise to express a view on whether they were 15 suitable or whether anyone designing the 16 searches, who did not already know that 17 Appendix 6 existed and the nature of its 18 content, reasonably could have selected 19 additional search terms that would have returned 20 that document." 21 So you're essentially making the same point 22 that we saw earlier in your paragraph 64 but in 23 a less direct way. Are you essentially asking 24 a form of rhetorical question there, with 25 a pregnant suggestion, that no one could 100 1 reasonably have selected a search term to pick 2 Appendix 6 up, without knowing the existence of 3 the document beforehand? 4 A. I'm saying I don't know whether such search 5 terms could have been devised to do that, but 6 the information I've seen or the information 7 I've been provided with is that there's nothing 8 within the face of that document that would 9 immediately lend itself to being identified by 10 the search terms that would have normally been 11 used in such circumstances. 12 Q. Is that how the Post Office thinks that search 13 terms work, that it's all about the design of 14 the search terms. If search terms are 15 reasonably selected and settled upon, they will 16 likely turn up the relevant documents? 17 A. Yes, except that search terms is not the only 18 device that's used to try to identify the right 19 documents. 20 Q. What if I suggest to you that search terms are 21 merely part or the start of an exploration for 22 relevant material in which search terms may lead 23 a reviewer to a relevant document? The document 24 itself needs to be read to see whether it is 25 associated with other documents which may also 101 1 be relevant but are non-responsive to the search 2 terms? 3 A. Yes, I think you've described it more eloquently 4 than I had but I think it's the same point that 5 I just made: it's not the start and finish. 6 Q. So it's a train of enquiry -- 7 A. (The witness nodded) 8 Q. -- starting with search terms but not ending 9 with search terms? 10 A. Well, that the train of enquiry starts with 11 identifying the relevant custodians where the 12 data may be held, et cetera, before the search 13 terms can be applied. 14 Q. But, essentially, I think you're agreeing with 15 me that the use of search terms can only be part 16 of a disclosure strategy. They will lead you on 17 a train of enquiry and the train must be 18 followed? 19 A. Yes. 20 Q. Can we turn, then, to the Post Office's approach 21 to reviewing family documents, and can we turn 22 to page 22 of your witness statement. Can we 23 read together paragraphs 67 to 70. You say: 24 "I have been informed by HSF and [Peters & 25 Peters] that, across the Inquiry's disclosure 102 1 requests, various approaches have been taken to 2 reviewing family documents. 3 "Depending on the nature of the relevant 4 disclosure request and what HSF or [Peters & 5 Peters] considered to be reasonable in the 6 circumstances, in some instances HSF or [Peters 7 & Peters] reviewed whole families of documents 8 ... and assessed whether each (or all) should be 9 produced to the Inquiry. 10 "However, in other cases, HSF or [Peters & 11 Peters] only reviewed the documents that were 12 returned by the search terms ... 13 "[The Post Office] was not involved in the 14 operational decisions about review approaches 15 for family documents. These were taken under 16 HSF's and [Peters & Peters'] general authority 17 to progress disclosure." 18 A. Yes. 19 Q. So, again, just to summarise all the threads 20 together, different approaches were taken by 21 your advisers to different requests, as to 22 whether they reviewed a family of documents or 23 not, or whether they just looked at the document 24 that was a hit, was responsive to the search 25 term? 103 1 A. Yes. 2 Q. The choice was made by your advisers according 3 to what was considered by them to be reasonable? 4 A. Yes. 5 Q. The Post Office was not involved at all, such 6 decisions were taken under a general authority 7 given to the lawyers? 8 A. Yes. 9 Q. So, until this whole issue blew up, neither you 10 nor, to your knowledge, anyone else in the Post 11 Office, knew what was going on in this regard? 12 A. Correct. 13 Q. Is your understanding that this approach adopted 14 by the lawyers was done on a Rule 9 by Rule 9 15 basis, ie "For this Rule 9, we are going to 16 review families, but for this Rule 9, we're not 17 going to review families", which is how your 18 witness statement reads? 19 A. I'm not sure I would see it as being done just 20 at the level of the Rule 9. I think it would 21 depend on the nature of the investigations in 22 relation to each Rule 9. So you would start 23 your investigations as we have just discussed 24 and, depending on where that took you, might 25 lead to approach, so rather than getting the 104 1 Rule 9 at the start and saying, "We just won't 2 look at family documents", I don't think there 3 was ever any question of that. 4 Q. So there might be different approaches to 5 reviewing families within a Rule 9? 6 A. Yes. 7 Q. So "For this bit, questions 1 and to 20, we will 8 review families but for questions 21 to 40, we 9 won't". 10 A. I'm not sure it's as black and white as that, 11 even at that level either. I think it would 12 depend on the nature of the investigations that 13 were ongoing and the way responsive documents 14 were being identified and, at that level, 15 decisions would then be taken. But I'm sorry, 16 I'm not familiar with the detail. 17 Q. Can we just look at paragraph 115 of your 18 witness statement on page 36. Again, this is 19 answering the "Who's responsible" question and 20 you're dealing in this part of the witness 21 statement with family documents. You say: 22 "As set out above in response to Request 1, 23 HSF did not require or instruct its reviewers to 24 review family documents (although reviewers had 25 the functionality to review family documents and 105 1 so could do so where they considered it would 2 assist them). To the extent that documents were 3 not identified and produced to the Inquiry 4 because of that decision, primary responsibility 5 logically sits with [Herbert Smith Freehills]." 6 This reads slightly differently, that your 7 understanding is there was no general 8 instruction to review family documents but there 9 was a function available to a reviewer and the 10 individual reviewer could decide to review the 11 family or not. Is that your understanding of 12 what went on? 13 A. So I think the -- my understanding is that the 14 individual reviewer had the capability, 15 something to do with the reviewing pane and how 16 it might have been shown. 17 Q. So a panel on the screen? 18 A. A panel on the screen. But there would have 19 been a prior discussion as to whether or not, as 20 a sort of a starting point, the family documents 21 should be looked at for that particular aspect, 22 which was presumably based on discussions 23 internally on what was being seen and advice 24 from more senior colleagues as to whether that 25 was an appropriate course of action. 106 1 Q. So it goes a bit beyond what's mentioned in 2 paragraph 115 here, which reads as if it was up 3 to the individual reviewer to decide? 4 A. I think that's probably unfair, yes. I think 5 there was -- there would have been discussions 6 going on but, ultimately, the documents were 7 there and could have been looked at if the 8 course of enquiry had suggested to that 9 individual reviewer that they should be looked 10 at. 11 Q. To your knowledge, were there any standing 12 instructions about when and in what 13 circumstances family documents were or were not 14 to be looked at by a reviewer? 15 A. I'm not aware of any. 16 Q. Would you agree that such standing instructions 17 would engender consistency across the disclosure 18 exercise? 19 A. Yes, provided they also allowed for discretion 20 where that was needed in relation to 21 a particular request. 22 Q. They would allow somebody looking back after the 23 event to see the rationale recorded for when 24 families were to be reviewed and when they were 25 not? 107 1 A. Which seems to me a different point as to -- 2 I mean, there's a difference between a general 3 guidance and then someone recording the 4 rationale, and I'd see those as two distinct 5 steps. 6 Q. Yes. Sticking with the guidance then, you're 7 right to pick me up on it, instructions to 8 reviewers that say, "These are the benefits of 9 looking at families of documents, you shouldn't 10 just think that because you're looking at a hit, 11 a responsive search term document, that's the 12 end of the exercise". 13 A. Yes. 14 Q. You should consider, by reading the document in 15 detail to see whether it refers in its -- the 16 body of the document to other documents, you 17 should check to see whether it's an attachment 18 to an email and go back to the parent email. 19 You should see whether it is part of a file, 20 a zip file, for example, and see whether any of 21 the other documents are relevant to the Rule 9 22 that you're considering. Was there anything of 23 that nature? 24 A. I'm not aware. 25 Q. Is there now? 108 1 A. Not -- again, I'm not aware of any. 2 Q. Can we go back to paragraph 68, please, which is 3 on page 22. You say: 4 "Depending on the nature of the relevant 5 disclosure request and what [Herbert Smith 6 Freehills] or [Peters & Peters] considered to be 7 reasonable ... in some instances [they] reviewed 8 whole families ..." 9 Yes? 10 A. That's the information I've been provided with, 11 yes. 12 Q. So whether families were or were not looked at, 13 was all down to what the lawyers considered 14 reasonable in the circumstances? 15 A. Yes. 16 Q. What informed the issue of reasonableness? 17 A. I don't know what they considered. 18 Q. Well, was it the number of documents that you 19 might have to look at if you decided to have to 20 look at the parent email, the costs of doing it, 21 the time it might take? 22 A. Yes, all of which are normal considerations in 23 any disclosure exercise to consider the range of 24 documents to be considered in the time available 25 that's going to produce the most relevant 109 1 documents. 2 Q. But what were they told, to your knowledge? 3 A. By Post Office? 4 Q. You've, as part of the remediation exercise, 5 I think taken a look backwards to see what was 6 going on. What was going on in relation to 7 family documents? You've explained it in 8 general terms here in paragraph 68 -- it all 9 depended on what the lawyers considered to be 10 reasonable -- I'm asking for some more detail on 11 that, please? 12 A. I'm sorry, I don't have any more detail. 13 Q. So if we took Appendices 3 and 6 as examples, 14 Appendix 3 was responsive to a search term, 15 presumably because it had the word "policy" in 16 it -- 17 A. Mm-hm. 18 Q. -- and "Investigation" and "prosecution" in it? 19 That appendix, Appendix 3, was part of a family 20 of documents, in fact many families of 21 documents -- 22 A. Yes. 23 Q. -- across the archive. It was often sent, as 24 I've said, as a zip file as part of a series of 25 appendices. So despite that document being part 110 1 of a zip file or within a suite of other policy 2 documents and the whole set of appendices 3 constituting the procedure which investigators 4 were to apply, because a decision had been made 5 not to look at families in this instance, 6 a reviewer would simply not look at those other 7 documents; is that right? 8 A. That's my understanding. 9 Q. I've got a hit, it's got the word "policy" and 10 "investigate" and "prosecute" in it, that's the 11 end of the matter. I don't see where this 12 document came from and whether it's part of a -- 13 or what's in the rest of the family; I just move 14 on to the next document? 15 A. I just don't feel able to comment on that 16 further, I'm afraid. 17 Q. Can we turn to paragraph 115 of your witness 18 statement, please. You deal with -- sorry, 19 we've looked at 115. 20 In terms of remediation, you tell us, 21 I think, that a decision has been now taken to 22 review all family documents; is that right? 23 A. That was done in relation to -- as part of the 24 remediation exercise to everything that had been 25 disclosed up to now, yes. 111 1 Q. Why was the decision taken to review all family 2 documents if previously it was thought that it 3 was unreasonable to have to do so? 4 A. Such was the level of concern that I think we 5 just felt we couldn't take the risk that 6 something else might have been missed and so we 7 wanted to go back and make absolutely certain as 8 to whether there was anything else relevant. 9 Q. Can I turn to de-duplication, please, and go to 10 pages 24 and 25 of your witness statement. If 11 we can read 78 at the bottom. Just by way of 12 explanation, at 78 you say: 13 "There are ... 2 relevant levels at which 14 de-duplication can be applied: 15 (a) family-level de-duplication -- 16 duplication is analysed as between complete 17 families of documents, excluding duplicate 18 families and ensuring only 1 copy of a family 19 survives computer processing to be manually 20 reviewed or produced ..." 21 Then secondly at (b): 22 "item-level de-duplication -- only 1 copy of 23 a document survives computer processing to be 24 manually reviewed or produced, even if it 25 appears in different family contexts." 112 1 Just stopping there, so that we understand 2 it in a simple example, if I've got one email 3 with five attachments to it and I've got another 4 email, which is in exactly the same terms, with 5 the same five attachments to it, family level 6 de-duplication will de-duplicate that? 7 A. Yes. 8 Q. If I have one email with five attachments to it 9 and another email with seven attachments to it, 10 one of which is the same attachment as one of 11 the attachments to the first email, item-level 12 de-duplication will de-duplicate the second 13 family as well? 14 A. If both emails were exactly the same, yes. 15 Q. Yes. So if both attachments were exactly the 16 same? 17 A. Yes. 18 Q. Thank you. You say: 19 "I now understand that item-level 20 de-duplication is rarely used in eDisclosure 21 because it can have significant unpredictable 22 and potentially undesirable impacts on document 23 review." 24 Who told you this? 25 A. I believe it came out through discussions with 113 1 Herbert Smith Freehills. 2 Q. Why might item-level de-duplication have 3 significant consequences and therefore is rarely 4 used? 5 A. Because it could result in family documents not 6 being available for a reviewer to see, so that 7 reviewing pane we spoke about earlier, they 8 wouldn't appear there. 9 Q. So it hides from the reviewer's view, not only 10 the single document, which is the duplicate of 11 the document he or she is looking at, but it 12 hides from the reviewer's view the other 13 documents within the family? 14 A. That's my understanding, yes. 15 Q. Why might it have unpredictable consequences? 16 A. Well, it would depend on the nature of the 17 document as to precisely what was lost and you 18 wouldn't know that in advance. 19 Q. Why might there be undesirable consequences of 20 this type of item level de-duplication? 21 A. Because it removes the ability of the reviewer 22 to see the relevant documents. 23 Q. Is it your understanding that the approach that 24 was, in fact, taken, the item-level 25 de-duplication, on occasions, meant that 114 1 a reviewer would not see the other documents in 2 the family, but also that the other documents in 3 the family were excluded from other keyword 4 searches being undertaken as part of the 5 disclosure exercise? 6 A. Yes, I believe that's correct. 7 Q. So this non-standard approach, which is rarely 8 used, and which has a series of significant 9 undesirable and unpredictable consequences, was 10 used, but all without the Post Office's 11 knowledge? 12 A. Yes. 13 Q. Aside from who, as between Herbert Smith 14 Freehills and Peters & Peters on the one hand 15 and KPMG on the other, was responsible for that 16 decision, do you know why, in this Inquiry, 17 which is itself investigating non-disclosure and 18 indeed, in the very Rule 9 we're looking at, was 19 searching for documents about the Post Office's 20 policy of giving or withholding documents in 21 investigations and proceedings, this 22 non-standard, rarely used approach was in fact 23 used? 24 A. No. 25 Q. Presumably when it was discovered, it caused 115 1 more than a minor ruffling of feathers? 2 A. Yes. 3 Q. Do you now know who was responsible for it? 4 A. No, it appears to be either a miscommunication 5 by Herbert Smith Freehills or a misunderstanding 6 by KPMG. I think both appear to have had 7 different understandings of what was being 8 instructed or what was being asked of them, and 9 I haven't, to date, been able to understand it 10 beyond that. 11 Q. Can we look at paragraph 80 which is further 12 down the page, please. You say: 13 "I understand that instructions to 14 de-duplicate were given to KPMG by [Herbert 15 Smith Freehills] and [Peters & Peters] and were 16 implemented by KPMG. I understand from [Herbert 17 Smith Freehills] and [Peters & Peters] that they 18 did not intend item-level de-duplication to be 19 applied in connection with the document searches 20 associated with responding to Rule 9 requests 11 21 and 14 (or at all). My current understanding is 22 that item-level de-duplication was applied as 23 the consequence of either a miscommunication by 24 [HSF/Peters & Peters] (who have told me they 25 didn't intend it to be applied) or 116 1 a misunderstanding by KPMG (who have explained 2 that they understood that their instructions 3 were to apply item-level de-duplication)." 4 So the lawyers say they didn't intend and 5 didn't instruct item-level de-duplication to be 6 used in any Rule 9 searches, yes? 7 A. I'm just pausing as to whether it's in any. 8 There may have been some in which it was 9 appropriate, I don't know, but my understanding 10 is it wasn't instructed in relation to these 11 Rule 9s. 12 Q. It was the addition of the words "or at all" in 13 the fifth line -- 14 A. Okay. 15 Q. -- that led me to ask that question, because 16 your witness statement speaks of, essentially, 17 all Rule 9 requests, not just 11 and 14. 18 A. Yes. In which case, that would have been my 19 understanding. Yeah. 20 Q. So the lawyers say they didn't intend item-level 21 de-duplication to be used in any Rule 9 Request, 22 and KPMG say they understood their instructions 23 to be to apply item-level de-duplication? 24 A. Yes, that's what I've been told. 25 Q. Isn't this kind of unattractive squabble between 117 1 highly paid city advisers why it is essential 2 for the client itself to be involved to 3 a sufficient degree to ensure an understanding 4 of what's being done in its name? 5 A. Yes, though whether Post Office being actively 6 involved in discussions would have enabled us to 7 understand that a misunderstanding had occurred, 8 I don't know. 9 Q. But by subcontracting the disclosure issue out, 10 you run exactly the kind of risk that has played 11 out in this case, don't you? By "you", I mean 12 the Post Office. 13 A. Yes. 14 Q. Can I turn, then, to remediation and, by 15 "remediation", I mean correcting past mistakes 16 and ensuring that other, similar mistakes do not 17 happen again. 18 In relation to the family issue, review of 19 family documents issue, as we've established 20 already, the remediation in relation to family 21 documents is essentially complete because, as 22 you've explained, a decision has now been taken 23 to review all family documents. 24 A. A decision was taken to review all the family 25 documents that had previously been produced. So 118 1 going forward, I think it will still be taken on 2 a case-by-case basis. 3 Q. That's what I was about to ask you. Is it in 4 relation to all new Rule 9 requests that family 5 documents will be reviewed, or is it in relation 6 to past Rule 9s and, if past Rule 9s, which and 7 why? 8 A. So the decision was taken to remediate in 9 relation to all previous ones out of 10 an abundance of caution, because of the level of 11 concern. In terms of as we move forward, there 12 will be much closer analysis as to how those 13 decisions are taken as to when to review family 14 documents. But my understanding is that the way 15 it could multiply the amount of documents that 16 need to be reviewed, it still needs to be looked 17 at on an individual, case-by-case basis, as to 18 whether or not that's the appropriate outcome. 19 Q. Again, as we sit here today, has any guidance 20 been developed to tell the reviewers when to 21 look at a family and when not to look at 22 a family? 23 A. I'm not aware of any specific guidance. 24 Q. So, in relation to Rule 9s that the Inquiry is 25 presently serving, there doesn't presently exist 119 1 any guidance to reviewers about when they should 2 look at a family of documents and when they 3 shouldn't? 4 A. I'm not saying it doesn't exist, I'm saying I'm 5 not personally aware. My team are certainly 6 more -- much more actively involved in 7 discussions in relation to current requests from 8 the Inquiry. I'm sorry, I don't know whether 9 they may have seen any. 10 Q. In relation to the search terms issue, can you 11 explain briefly what has been done in relation 12 to remediation of that problem? 13 A. So in relation to the specific issue that was 14 identified, which I understand was in relation 15 to the -- I think the failure to shorten the 16 word "guidance" to "guide", that was done and 17 the remediation of that is complete. There's 18 been a more broad exercise in terms of -- 19 I regard it more as quality assurance rather 20 than it necessarily being a mistake, but to 21 whether or not, on the basis of all the 22 additional information that has been accrued up 23 until now, different search terms should now be 24 gone back and applied and that's the extensive 25 work which has been undertaken. 120 1 Q. Is that part of what you describe in your 2 witness statement -- we needn't go there, it's 3 108 and following -- as part of the structural 4 review or does it sit outside that? 5 A. It sits outside that. 6 Q. Who is undertaking that work in relation to 7 search terms? 8 A. Herbert Smiths and Peters & Peters. 9 Q. Whilst I touch upon it, can you explain in 10 summary terms what the structural review is, 11 please? 12 A. Sure, if you just give me a moment to refresh my 13 mind. 14 Q. It's 108 on page 34. 15 A. Thank you, yes. I know I've made a note for 16 myself but I can't immediately find it. Thank 17 you, yes. So it's centred around the use of the 18 electronic discovery reference model, which 19 I have annexed to my statement, and it's 20 checking, to the best level of achievable, 21 whether all relevant documents that make up 22 disclosure -- so it's checking that to see 23 what's been collected already, what still needs 24 to be collected, how it's held on Relativity, 25 whether there's any potential to restructure the 121 1 Relativity system to enable better search 2 functions. 3 It's looking at how we might increase 4 technological options in Relativity, which might 5 allow, for example, some computer assisted 6 searches to be done and it's also looking at all 7 the steps in the process. So what's done by 8 whom and looking at how we might reduce risks 9 and make achievable improvements. 10 Q. As part of that, you refer to the fact that the 11 Public Inquiry Team is presently advertising for 12 a disclosure specialist lawyer who would engage 13 in the granularity of the disclosure processes? 14 A. Yes. 15 Q. Does it follow that there has not been a person 16 within the Post Office who has been engaged with 17 the granularity of the disclosure processes? 18 A. So the lawyers that are in the team have 19 experience of either public inquiries or similar 20 work but I don't believe they have specific 21 expertise in disclosure matters in the way that 22 I think I need in the team. 23 Q. That disclosure specialist lawyer hasn't been 24 appointed yet; is that right? 25 A. So in the -- no, it hasn't. So we haven't 122 1 appointed someone as yet. In the interim, what 2 we have had is two senior lawyers from Burges 3 Salmon and Fieldfisher who have worked 4 effectively in-house as part of the team since 5 June to help us provide that extra challenge 6 function, and the idea is that the recruitment 7 permanently will take the place of those 8 lawyers. 9 Q. Is that new post, the disclosure specialist 10 lawyer, as has been reported in the media, to be 11 paid, in part, by reference to a bonus? 12 A. Yes. 13 Q. What is the bonus metric? 14 A. So Post Office, like many organisations, runs 15 a bonus scheme to which its senior manager 16 professionals and management are entitled to 17 participate in, which has business-wide 18 objectives, which is set for the whole 19 organisation, including things like financial 20 targets. The team is entitled to take part in 21 that in the same way that other parts of Post 22 Office are. There are -- in the current scheme 23 and in any future schemes, there are no metrics 24 related directly to the Inquiry. 25 Q. So that lawyer and other lawyers, is this right, 123 1 are not being paid bonuses that are related to 2 their performance in Inquiry work? 3 A. So I think we have to look at it at two levels. 4 So, first of all, there's a decision which is 5 taken by the remuneration committee as to 6 whether or not the corporate-wide objectives 7 have been met and that triggers the entitlement 8 in principle to payment of a bonus. At a team 9 level, as you'd expect, there are personal 10 objectives for each member of the team, which 11 are focused on, in the Inquiry Team, support for 12 the Inquiry. Their performance is then looked 13 at in the round at the end of the year, looking 14 at what they've delivered and in what context, 15 with what standards of behaviour, et cetera. 16 Higher levels of performance could lead to the 17 achievement of a higher bonus award. But the 18 decision has to be taken in the first place that 19 it's payable at all. 20 Q. Lastly, and in less than five minutes, can we 21 look at the hard-copy document review. This is 22 a discrete issue in the sense that it's not 23 related to the three disclosure issues 24 identified by the Chair in his previous 25 decisions. That is the hard copy document 124 1 review being undertaken by the Post Office and 2 its advisers. Can you explain to us in summary 3 form what the hard-copy document review is? 4 A. So there was, I think after concerns had been 5 raised previously in the way that -- some of 6 what Fintan has described about the slightly 7 haphazard way that data had been stored, there 8 was a review undertaken of where all data, 9 hard-copy data, was stored. That was then 10 audited by an external firm, which reviewed all 11 the sites to see what was there. That's led to 12 the discovery of some additional hard-copy 13 material which has been produced to the Inquiry. 14 It also led to the discovery of some additional 15 electronic material such as back-up tapes, where 16 some of the work is still ongoing. 17 Q. Can we look, please, at POL00124517. This is 18 a letter from Herbert Smith Freehills to the 19 Inquiry of 22 August that deals with the hard 20 copy document review and, under the heading 21 "Back-up tapes", if we scroll down, please, the 22 author says: 23 "As set out [in a previous document] in 24 October/November 2002, [the Post Office] 25 identified 42 ..." 125 1 Just stopping there, I think that's 2 subsequently corrected to 37. 3 A. Correct, yes. 4 Q. "... [37] back-up tapes with unknown contents at 5 its Chesterfield site." 6 So this is telling us that nearly a year ago 7 now, 37 back-up tapes were found in 8 Chesterfield. 9 A. Correct. 10 Q. If we go over the page, please, at paragraph 7: 11 "... in December ... and again in February 12 [the Post Office] instructed KPMG to prepare 13 indices of data stored on the tapes ..." 14 In paragraph 8 KPMG prepared a total of 91 15 indices. 16 In paragraph 9: 17 "... following ... investigation of the 18 files, POL decided to process the files 19 remembered to in [some of the] indices ... onto 20 a separate, dedicated, Relativity workspace for 21 further examination." 22 Is that what you referred to earlier, 23 separating something out for a smaller scale 24 review? 25 A. Yes. 126 1 Q. Paragraph 10: 2 "Processing the data onto the Relativity 3 workplace has taken much longer than 4 anticipated." 5 Paragraph 11: 6 "De-duplication, at the 'parent' document 7 level, has been applied at the processing stage 8 across the 19 datasets ... There is very 9 significant duplication as between the 10 datasets/sessions ..." 11 A reduction from 19 million documents to 12 4.8 million documents. Over the page, please, 13 at 13: 14 "POL recognises the importance of continuing 15 to assess the likelihood of new Inquiry relevant 16 documents being backed up on the 21 17 datasets/sessions. [Your] current plan is as 18 follows ..." 19 That's set out in the remainder of 20 paragraph 13. 21 Then there's an update, please. Can we look 22 at POL00126338. This is a letter that came last 23 Thursday under the heading "Back-up tapes". 24 Origin of the back-up tapes, as we know, they 25 were found in Chesterfield. 127 1 If we go to page 3 of the letter, please. 2 At the top of the page -- I should say on the 3 previous page there's a summary of some of the 4 steps that were set out in the previous letter. 5 A. Yes. 6 Q. "Whilst POL is not yet able to confirm with 7 certainty how long these steps will take, POL 8 currently estimates it may take between one to 9 two months before POL has identified a potential 10 review pool. This will be subject to required 11 machine time which will be dependent on the 12 eventual data sizes which need to be migrated 13 and de-duplicated." 14 Then search terms will be applied. It will 15 take approximately two days to run the search 16 terms and KPMG say it will take approximately 17 two weeks for the responsive documents to be 18 migrated to the Inquiry platform. 19 13: 20 "The migrated documents will then need to be 21 de-duplicated." 22 That's essentially de-duplicated against 23 what you've already disclosed to the Inquiry. 24 A. Yes. 25 Q. 14: 128 1 "... POL will then assess the data that is 2 left ... and consider ... targeted searches ..." 3 15: 4 "... given the large amount of data from 5 these 21 datasets, further consideration will 6 have to be made to ensure any review is 7 proportionate ... [The Post Office] will keep 8 the Inquiry informed ..." 9 A. Correct. 10 Q. So essentially, it's this: it's going to be 11 another one to two months before the data that 12 was discovered last October/November, is even 13 assembled into a pool that can be searched? 14 A. That's correct but my understanding is that the 15 steps that have been taken have been essential 16 in order to get the data down to even any kind 17 of manageable size so that that review could 18 take place. 19 Q. Has there been any assessment as to the 20 likelihood that the data that has not yet been 21 reviewed is of relevance to the three phases of 22 the Inquiry that have already been completed and 23 Phase 4, that we are part way through at the 24 moment? 25 A. So, as a starting point, my understanding is 129 1 that, by their very nature, as back-up tapes, 2 the expectation is that they won't contain new 3 information. But I don't understand us to be at 4 the stage of being able to work out what may or 5 may not be relevant. 6 Q. Is that because there isn't an understanding of 7 what they are a back-up of? 8 A. Yes, correct. 9 Q. Finally, on the remediation exercise, has the 10 Post Office got a target date which it can now 11 inform the Inquiry of completion? 12 A. So my understanding is the vast majority of the 13 work has been done. There are some further 14 target dates over the course of September, so 15 14 September, some more on 29 September, but 16 then an overall expectation that any remaining 17 work will have been achieved in good time before 18 hearing dates are set for the criminal case 19 studies work. 20 MR BEER: Mrs Wills, thank you very much. 21 THE WITNESS: Thank you. 22 MR BEER: Sir, apologies for going slightly into the 23 lunch break. I wonder if you have any questions 24 or if not, might we break until 2.05. 25 SIR WYN WILLIAMS: I don't have any questions so we 130 1 will break until 2.05. 2 Mrs Wills, thank you very much for your 3 detailed statement and the detailed answers to 4 the many questions put to you. 5 THE WITNESS: Thank you. 6 MR BEER: Thank you very much. 7 (1.05 pm) 8 (The Short Adjournment) 9 (2.12 pm) 10 MR BLAKE: Good afternoon sir, can you see and hear 11 me? 12 SIR WYN WILLIAMS: Yes, I can thank you very much. 13 MR BLAKE: Sorry for the delay, sir. We have been 14 having a little difficulty with the transcript. 15 It is, I believe, now fixed but, if there is 16 a difficulty with the transcript during this 17 afternoon's evidence, we will just continue and 18 then a transcript will be available at the end 19 of the day. 20 SIR WYN WILLIAMS: Yes, that's fine, Mr Blake. 21 MR BLAKE: Thank you. Can I call Mr Tombleson, 22 please. 23 PAUL JOHN TOMBLESON (sworn) 24 Questioned by MR BLAKE 25 MR BLAKE: Thank you, can you give your full name, 131 1 please. 2 A. Paul John Tombleson. 3 Q. You should have in front of you a witness 4 statement, it is dated 22 August 2023; is that 5 correct? 6 A. That's correct. 7 Q. Can I ask you to look at page 14 of that 8 statement. Is that your signature at the bottom 9 there? 10 A. Yes, it is. 11 Q. Can you confirm that statement is true to the 12 best of your knowledge and belief? 13 A. Yes, I can confirm that. 14 Q. Thank you very much, Mr Tombleson. That 15 statement which, for the purpose of the 16 transcript is WITN09960100 will be published by 17 the Inquiry and I'm just going to build on 18 what's already in that statement, rather than 19 ask you to set it all out again. 20 By way of a brief career history you are 21 a partner in the forensic practice at KPMG; is 22 that correct? 23 A. Yes, that's correct. 24 Q. You've been with KPMG for over 33 years; is that 25 right? 132 1 A. Since 1990, yes. 2 Q. You've been leading in eDisclosure since 2005? 3 A. That's correct. 4 Q. In your statement, you said you were engaged by 5 the Post Office from January 2021, so that's 6 during the non-statutory phase of the Inquiry; 7 is that right? 8 A. That's correct, yes. 9 Q. You're described as being engaged by the Post 10 Office. Who was or is your client: is it the 11 Post Office, is it Herbert Smith Freehills or 12 somebody else? 13 A. From a contractual perspective, it's the Post 14 Office. 15 Q. From a day-to-day relationship perspective? 16 A. From a day-to-day perspective, we took 17 instructions from the external solicitors in 18 relation to the various requests and responding 19 to those. 20 Q. Do you have standing instructions with regards 21 to disclosure, provided by either the Post 22 Office or Herbert Smiths or is disclosure 23 addressed on a more ad hoc basis? 24 A. We discuss with the external solicitors that 25 we're working with the various steps we take 133 1 throughout the disclosure process, the 2 eDiscovery process, starting from the collection 3 of data through to how it's processed, how it's 4 searched, how it's presented to them to review 5 and ultimately how it is produced. I think it 6 would be fair to say we don't have a single 7 document currently that we use as the basis for 8 that. We use internal standard procedures 9 within KPMG and, for every request that comes 10 in, we have a clear audit trail of discussions 11 and confirmations between us. 12 Q. When you say every request that comes in, we 13 know that the Inquiry issues Rule 9 Requests or 14 now it issues section 21 notices, are those 15 discussed on an individual basis as to how 16 they're approached, are they discussed on 17 a question-by-question basis or is there some 18 broader guidance in terms of, for example, 19 families, duplicates, et cetera? 20 A. I'd say they're discussed on 21 a request-by-request basis and sometimes on 22 a question-by-question basis within those 23 requests, if there are a lot of different 24 questions within the requests. 25 Q. I think it's the evidence we've already heard 134 1 today that there is no standard approach to, for 2 example, what to do about family documents, what 3 to do about duplicate documents. Those are on 4 a request-by-request basis? 5 A. Certainly the de-duplication, that is 6 an eDisclosure process that we discuss on 7 a request-by-request basis. I can't really 8 comment on the family review side of things. 9 Q. Why is that? 10 A. Because we're not doing the review. I can 11 comment on how we present the documents for 12 review, which would be that we would present 13 only the responsive documents for review but 14 I think your question was a bit broader than 15 that, whether we've got guidance for the family 16 document review and we don't. 17 Q. Do you have sight of, for example, Herbert Smith 18 Freehills' own internal guidance as to what to 19 do with particular documents? 20 A. No, I don't. 21 Q. We've heard, before the lunch break, about cost 22 pressures. Were you aware of any cost 23 pressures? 24 A. Yes, one of my roles as the partner responsible 25 for KPMG's work is to have a regular dialogue 135 1 with the Post Office around things that could be 2 done to manage costs without compromising the 3 quality of the service we provide. 4 Q. Mrs Wills, for example, talked about resourcing 5 concerns that have been raised with KPMG. Have 6 you ever been given the impression that there is 7 insufficient funding to carry out 8 a comprehensive disclosure exercise? 9 A. No, never been given that impression. 10 Q. We heard about changes between January and May 11 2023, some work being carried out in India; can 12 you please tell us a little bit about that? 13 A. Yes, so one of the -- the conversations started 14 around cost rather than resourcing. One of the 15 advantages of a firm like ours is that we have 16 a dedicated team in India, an offshore team. 17 These are forensic technology eDisclosure 18 professionals who work, for all intents and 19 purposes, as part of our KPMG UK team, they're 20 a dedicated team that works with us and, in 21 certain circumstances, we're able to bring them 22 on to undertake tasks in the same way as we 23 would with some of our UK resources. 24 So the conversations around that started -- 25 may have been earlier this year, and the Post 136 1 Office gave us security clearance to bring those 2 individuals on in around March and April time. 3 So they undertake some of the activity around 4 searching, moving data between workspaces but 5 always with a level of quality control that's 6 provided by a member of our senior onshore team. 7 Q. Are you aware of any document review taking 8 place outside of this jurisdiction? 9 A. No, I'm not. 10 Q. We've heard that you were involved in the 11 non-statutory phase and then in the statutory 12 phase. Mr Canavan's evidence this morning 13 suggested there was a lack of understanding at 14 the Post Office with regards to what a statutory 15 Inquiry might involve. Were you aware of that 16 at all? 17 A. No, I wasn't, no. 18 Q. Do you feel you were sufficiently resourced 19 throughout the lifetime of this Inquiry? 20 A. Mean I would say the level of activity has 21 certainly picked up in the last few months and 22 Mrs Wills referred to a conversation we had 23 where I think we did start to increase the size 24 of our team at that point in time because there 25 was numerous requests. Just by indication, 137 1 we -- our team was about 15 strong in May and is 2 close to 30 strong at the moment but that's, in 3 large part, to also support the work on the 4 remediation. 5 Q. So it's doubled since May. If we look back to 6 the summer of 2021, how many people made up your 7 team? 8 A. Probably six. Six or seven. 9 Q. Over what period did it increase from six to 10 double that? 11 A. Well, the workspace for the Inquiry was set up 12 in October 2021 and I would have thought 13 November/December '21 was when it really started 14 to increase in size. 15 Q. I want to ask you about keyword searches. 16 You've explained that your instructions came 17 from Herbert Smith Freehills and not from the 18 Post Office. As someone with significant 19 experience in eDiscovery matters, do you think 20 it's a good idea for law firms to closely liaise 21 with their clients in respect of identifying 22 search criteria? 23 A. I don't know how to answer that question, 24 really. I mean, in terms of our role, our role 25 was to receive the search terms from Herbert 138 1 Smiths as the RLR. I wasn't aware of the 2 process before we received them that those 3 search terms had gone through, whether there had 4 been a consultation process with the Post Office 5 or not. 6 Q. You may not have been aware in this particular 7 case but you have been involved in a number of 8 significant pieces of litigation, public 9 inquiries, perhaps. Are you aware from other 10 experience that you have, of there being visibly 11 closer liaison between the law firm and the 12 client than there is perhaps in this particular 13 case? 14 A. I think it would only be fair to say I've seen 15 it happen both ways. I've seen it with external 16 solicitors taking the primary role or more 17 consultation with the client. It's difficult to 18 give a definitive answer to that, I'm afraid. 19 Q. In your view, which has worked better? 20 A. I think it really depends on the subject matter 21 of the case and where the knowledge and 22 experience in order to structure the search 23 terms resides. 24 Q. Given your experience in this particular case, 25 do you think that there was insufficient liaison 139 1 with the lay client, from what you've seen? 2 A. As I said before, I haven't seen, but my answer 3 to that would be, given the subject matter, 4 I think there should have been liaison but 5 I don't know what liaison there had been. 6 Q. Let's take the prosecution ID codes document as 7 an example. 8 A. Yeah. 9 Q. Do you see it as important to liaise with 10 a client in the kind of cases that we've seen 11 there, where the request is for a policy, 12 historic policies, that may not be searchable by 13 way of, for example, a keyword for "policy" 14 because they may have used, as Mr Beer 15 highlighted earlier, a different word, such as 16 a "guide"? Do you see it as important in those 17 kinds of circumstances for there to be close 18 liaison with the client? 19 A. That particular request, I didn't see the 20 wording of the request until relatively 21 recently. I saw the search criteria that were 22 constructed and the search terms that were 23 constructed but I didn't see the underlying 24 wording from the Rule 9 Request. 25 So I think I am answering your question by 140 1 saying I don't know whether the construction of 2 the search term was sufficient to find what the 3 Inquiry was looking for with that request. 4 Q. With your now large team of individuals working 5 for KPMG on this project, is there no one who 6 applies their mind to the wording of a Rule 9 to 7 the search terms that are applied or decided by 8 the law firm and whether those are appropriate 9 or not? 10 A. So our role is not to develop the search 11 criteria and the search terms themselves. Our 12 role is to help construct the syntax of those 13 terms, so they're as precise as possible and, in 14 the example you showed a little while ago with 15 truncated -- the use of a wild card character 16 and a truncated word, that's the sort of advice 17 that my team would give, rather than the -- does 18 it address the question in the underlying Rule 9 19 Request which, as I said, we don't see. 20 Q. Their role may not be to give that advice, but 21 surely people who work in your team might have 22 a view on whether keywords are appropriate in 23 particular circumstances. 24 A. Again, in this case, because we haven't seen the 25 underlying Rule 9 Request, I don't think we can 141 1 make a judgment on that. 2 Q. Have there been circumstances in this Inquiry 3 where your team at KPMG have seen a request or 4 have -- are aware of the contents of a request, 5 have seen the proposed search terms and have had 6 a view as to whether those search terms are or 7 are not appropriate? 8 A. Not appropriate. Whether they are generating 9 too many responsive documents or a high number 10 of responsive documents, that's the sort of 11 analysis that we're doing, which of the search 12 terms is generating the most hits, and then 13 a dialogue with the external solicitors and if 14 they want to revise those terms in any way. 15 Q. You may not express a view to Herbert Smith 16 Freehills, but are you aware of anybody within 17 your team ever having a view as to whether 18 search terms are effective or appropriate? 19 A. No, I'm not aware of that. 20 Q. I want to ask about family documents. At 21 paragraph 19 of your witness statement you say 22 that in the vast majority of cases the 23 instructions were to provide only responsive 24 documents and not their families. Depending on 25 the request -- and let's take the ID codes case 142 1 as an example -- can you see the benefit of 2 reviewing family documents? 3 A. Clearly, with hindsight, yes, in that particular 4 case. The balance is always one of a number of 5 documents to review within a very limited amount 6 of time and it's usually well recognised that 7 documents that are responsive to search terms 8 tend to have a higher relevance rate than those 9 which are not. 10 I'm not saying -- that doesn't mean you 11 shouldn't review family items but what I'm 12 saying is that there is a balance and decision 13 that needs to be taken about where to prioritise 14 the review effort. 15 Q. You're sitting there with your 33 years of 16 experience, significant experience in public 17 inquiry work or high-level litigation. If you 18 were told that the request is for historic 19 prosecutions policies and investigations 20 policies, dating back over a ten-year or more 21 period, what decision would you have been -- 22 would you personally have made in respect of the 23 reviewing of family documents? 24 A. I'm not a lawyer and I'm not conducting the RLR 25 role in this Inquiry. I can't answer that 143 1 without just purely speculating. 2 Q. You're not being asked to answer that as 3 a lawyer, but you're being asked to answer that 4 as someone who knows about data, about what 5 documents might or might not be responsive to 6 particular hits, the levels, numbers of 7 documents that you might receive. 8 A. Yeah. 9 Q. In those circumstances, if you're being asked 10 for, for example, policy documents, which may 11 have annexes, for example -- 12 A. Yeah. 13 Q. -- and you were being asked to give advice, not 14 in a legal context but just being asked whether 15 it would be helpful, more helpful, less helpful, 16 what would your position have been in respect of 17 family documents? 18 A. I think I would have started with a set of 19 search terms, I would have commenced the review 20 based on documents that are responsive to those 21 search terms and, if, through that review, other 22 things appeared relevant in terms of words which 23 weren't included in the search terms that we'd 24 applied upfront, then considered going back and 25 rerunning those terms also across the dataset. 144 1 Q. So is it fair to say that your approach would be 2 to review the documents, understand them, and 3 perhaps look further, depending on what you have 4 received? 5 A. Potentially. 6 Q. If families had been reviewed in this particular 7 case, what would have happened in respect of 8 that ID codes document? 9 A. That would have been identified in April 2022. 10 Q. I want to ask you about de-duplication. 11 Mrs Wills has said that item-level 12 de-duplication is rarely used in eDisclosure 13 because it can have significant unpredictable 14 and potentially undesirable impact on document 15 review; do you agree with that? 16 A. I agree with the first part of the document that 17 it's uncommon in eDisclosure. The set up of the 18 data and the structure of the workspaces in this 19 Inquiry is uncommon in so many ways. Typically, 20 you would have your de-duplication done at the 21 processing stage, right at the beginning, by 22 a single vendor. In this particular case, data 23 has arrived at various points from various 24 different matters, and it's been necessary to 25 consolidate that all in one place and then apply 145 1 de-duplication prior to review, so at a stage 2 you wouldn't ordinarily do de-duplication. 3 So a more nuanced solution has been 4 necessary. But I would still item-level 5 de-duplication is uncommon because it's 6 a workflow that is appropriate for a review of 7 responsive documents. 8 Q. The problem that you've identified in this 9 particular case, which you say required a more 10 nuanced solution, does that go back to the 11 evidence we heard from Mr Canavan about the Post 12 Office's repositories and historic data storage 13 and the discovery, for example, of further 14 repositories? 15 A. Not specifically. It's a more general point 16 that the data does not reside in one single 17 database. It resides in four very large 18 databases, each with tens of millions of 19 documents in them and so the disclosure process, 20 the eDisclosure process, requires us to run the 21 same searches in three or four different places, 22 then migrate that data into a single location, 23 and then de-duplicate in that location -- if 24 we're asked to, I should say. De-duplication 25 only happened when we were instructed to do so. 146 1 Q. But you've said it's rarely used and it occurred 2 because, in this particular case, it required 3 a more nuanced solution. Presumably, that is 4 quite a significant step and significant 5 decision to make then, a departure from the 6 usual position; do you agree with that? 7 A. Well, if the decision involved the discussion 8 and the clarification of the requirements, 9 involved the senior manager and the manager who 10 were the day-to-day leads on our work. 11 Q. So it required managerial approval on the KPMG 12 side of the work; is that correct? 13 A. Involved managers and senior managers on our 14 team understanding the requirements and making 15 sure we were clear on the requirements before we 16 implemented a solution, yes. 17 Q. How about on the Post Office or Herbert Smith 18 Freehills side of the fence? 19 A. So we -- my team was dealing with senior 20 associates on the Herbert Smiths side and not 21 dealing with the Post Office in relation to this 22 matter. 23 Q. So am I right to understand that this rarely 24 used solution, which was a nuanced solution to 25 a particular problem, was, on the Post Office's 147 1 side, dealt with only by senior associates at 2 Herbert Smith Freehills liaising on your side 3 with quite senior management? 4 A. Correct. 5 Q. Can you see a problem in that? 6 A. The team that we have working from -- I can only 7 comment from the KPMG side, but I've complete 8 confidence in the senior manager and the manager 9 on our side and subsequent senior members of my 10 team who have been working on this particular 11 workflow. 12 Q. But for this particular solution their 13 instructions came from people who are relatively 14 junior; do you agree with that? 15 A. I don't know what senior day-to-day -- our 16 senior associate within a law firm can cover, 17 you know, a wide range of experience, many years 18 post-qualified as well. My experience of 19 working with some of these people was that they, 20 you know, they were experienced and 21 knowledgeable lawyers who'd worked in the 22 eDisclosure space before. 23 Q. In your view, was this a careful decision that 24 was carefully recorded, for example? 25 A. There's an audit trail of the conversations that 148 1 happened and confirmation between us, as to what 2 the workflow would look like in order to execute 3 on this de-duplication strategy, yes. 4 Q. Can we look at page 7 of your witness statement. 5 There's a table on page 7. It's halfway down 6 the page. If we look at that table, can you 7 talk us through the problems that can occur, 8 using the ID codes as an example, in item-level 9 de-duplication in this particular case? 10 A. Yes. So this is a table which shows five 11 families of documents. In total there are 19 12 documents within this set of data. The emails 13 represent the parent documents, the attachments 14 represent the child documents to those families. 15 The way that family-level de-duplication 16 works -- sorry, before I say that, I should 17 point out that you can't see it in this image 18 here but Attachment 1, in this particular 19 example, is the document which is responsive to 20 the search terms and has been highlighted red -- 21 Q. So that is the policy document, I think we call 22 it, is it Annex 3 or -- 23 A. Correct, so that's the equivalent to Appendix 3. 24 Q. Appendix 3. 25 A. Under family-level de-duplication, one would 149 1 look at the families that the responsive 2 documents were in and de-duplicate any that were 3 identical. In this particular case, Family 4, 4 and Family 5 are identical and, therefore, one 5 of those two families would be de-duplicated and 6 one would remain in the review set. 7 Q. Pausing there, if we had family-level 8 de-duplication, the only document that would be 9 removed from view would be Family 5 because they 10 are identical to Family 4? 11 A. Yes, I mean, there's a way in which decisions 12 are made between whether it's 4 or 5, but 13 broadly that's correct. There would be 14 14 documents left in the dataset after family-level 15 de-duplication. 16 Q. Can you talk us through item-level 17 de-duplication? 18 A. So item level de-duplication, rather than 19 looking down the vertical, looks across the 20 horizontal, and it says there are five identical 21 responsive documents, which is Attachment 1, and 22 it de-duplicates that so only one version of 23 an attachment is presented to the reviewers and, 24 once again, there is a decision that needs to be 25 made about which one of those four families -- 150 1 or, sorry, five families -- is the master and 2 which one is the duplicates. 3 Q. Am I right to say, then, that if you were 4 carrying out item-level de-duplication on this 5 particular table, you wouldn't get Attachment 2, 6 Attachment 3, Attachment 4 and you wouldn't get 7 nearly every document in that table? 8 A. Yes, so the example I've given is where Family 1 9 has been selected as the master and that means 10 that you wouldn't see the attachments that 11 you've just named. 12 Q. So a decision to carry out item-level 13 de-duplication, we see the significance of such 14 a decision in that particular table; do you 15 agree with that? 16 A. That's correct. 17 Q. That is perhaps why you have carefully recorded 18 the decision that was made in that respect? 19 A. That's correct. 20 Q. Thank you. Can we have look at paragraph 80 of 21 Mrs Wills' statement, that's WITN09940200. It's 22 page 25 of that statement. Thank you, if we 23 scroll down, thank you. So Mrs Wills' evidence 24 is, she says as follows: 25 "I understand that instructions to 151 1 de-duplicate were given to KPMG by [Herbert 2 Smith Freehills] and [Peters & Peters] and were 3 implemented by KPMG. I understand from HSF and 4 P&P that they did not intend item-level 5 de-duplication to be applied in connection with 6 the document searches associated with responding 7 to Rule 9 requests 11 and 14 ..." 8 Those are the requests for the prosecution 9 policies, investigations policies: 10 "... (or at all). My current understanding 11 is that item-level de-duplication was applied as 12 the consequence of each a miscommunication by 13 [Herbert Smith Freehills/Peters & Peters] (who 14 have told me that they did not intend it to be 15 applied) or a misunderstanding by KPMG (who have 16 explained that they understood that their 17 instructions were to apply item-level 18 de-duplication)." 19 Now, if it was such a significant decision, 20 a nuanced solution to a particular problem, one 21 that was carefully recorded, how might it be 22 that the answer to that is not straightforward? 23 A. I think because -- I don't think my team or 24 I misunderstood what we were being asked to do. 25 And if I may -- maybe I'll refer to my table 152 1 again in a minute but, before I do that, we were 2 asked to provide statistics on the number of 3 hits that a search term had identified, in other 4 words how many responsive documents were there. 5 We weren't typically being asked to provide 6 statistics on the number of hits and their 7 associated families. The conversations that 8 would happen about de-duplication would then 9 start with, "This number of hits seems too 10 high", and the first set of conversations would 11 be around are there things around the search 12 terms that need refining, and the second set of 13 conversations would be about de-duplicating the 14 responsive documents, de-duplicating the hits, 15 so that the external solicitors only see one 16 version of a document. 17 And, as you can see from the worked example 18 here, the only way that works when you're doing 19 a review of only responsive documents is at 20 an item level, otherwise you leave lots of 21 versions of the same document in the dataset 22 which need to get reviewed time and again. 23 Q. You've said that the decision was recorded, what 24 does that record say, then? 25 A. The decision -- recorded in the communications 153 1 and the conversations between our teams, where 2 the requirements are set out, what they would 3 like, which is a reduction in the number of hits 4 by only giving us unique documents, one version 5 of a document for review, and our team 6 clarifying whether they would like us to 7 de-duplicate on the basis of an MD5# algorithm 8 just the responsive documents, that being 9 confirmed, and then the statistics then showing 10 that in a series of tables. 11 Q. So is your evidence that KPMG asked Herbert 12 Smith Freehills whether they wanted item-level 13 de-duplication and the answer that came back was 14 "Yes"? 15 A. That isn't -- no, that isn't my -- my statement 16 is that -- that expression, "item-level 17 de-duplication" was never really used until 18 March/April time this year. What we were 19 talking about was responsive document 20 de-duplication, which is the same thing. 21 Q. So you may not have used the words that I've 22 used, but you asked somebody at Herbert Smith 23 Freehills, presumably you know who it is, 24 presumably -- based on your previous earlier 25 evidence -- it was a senior associate at Herbert 154 1 Smith Freehills, for the significant decision, 2 "Do you want the responsive documents to be 3 de-duplicated?" and the answer that came back 4 was "Yes"? 5 A. That's correct and it wasn't one individual. We 6 dealt with a number of senior associates and 7 associates. Every single request and question 8 was dealt with separately. There was no 9 standard approach to de-duplication. 10 Q. So that was a question that you expected some 11 reasonable consideration to have been given to. 12 A. Yes. 13 Q. Was the question based on statistics? So would 14 you, for example, say, "If we don't de-duplicate 15 we'll have to review this number of documents, 16 if we do, we'll have to review this number of 17 documents", and the option was given to them as 18 to how many documents they wish to consider? 19 A. That's correct. Our role is to provide the 20 statistics pre-de-duplication and 21 post-de-duplication. 22 Q. Having considered Mrs Wills' evidence on this 23 point, do you understand why it may be suggested 24 that there was a miscommunication rather than 25 the simple answer that you have given? 155 1 A. Yes, I do understand that, yes. 2 Q. You understand why it may be put as 3 a miscommunication? 4 A. So I understand that we were very clear in what 5 we were doing, based on the instructions we were 6 given, the consequence of that is related to the 7 family items of duplicates of the documents that 8 were being reviewed. I don't know what the 9 intention of any of the lawyers was but, if 10 a review of family items of all versions of 11 documents was necessary, then there would be 12 consequences from this, clearly, as we've seen. 13 But I didn't see that documented in that level 14 of detail at that time. 15 Q. Do you see it as a problem that KPMG doesn't 16 apply its mind at all to the problem and only 17 provides statistics, an A or a B option, for 18 example? 19 A. I think you still apply in your mind in 20 delivering some of the technology solutions that 21 we are. I just felt like we have different 22 roles, and our role is, in this particular 23 Inquiry, not to interpret the wording of the 24 request and convert that into the wording that 25 needs to be then searched across the database. 156 1 Our role is the more technical side of 2 establishing which work spaces is the syntax 3 correct, and making sure from the quality 4 control perspective we're getting accurate 5 results. 6 Q. Do you think that KPMG should be, for example, 7 advising on the pros and cons of different 8 approaches? So the con in this case being -- or 9 the pro being, yes, you exclude a large number 10 of documents, the con being you might actually 11 need to read some of those documents? 12 A. I mean, I don't think we were asked for our view 13 on that. Again, with hindsight, if we went back 14 to March last year, would I have encouraged the 15 team to have a different conversation about the 16 full impact of both pros and cons? Then, yes, 17 probably I would. But, you know, I strongly 18 believe, looking at the wording of the request 19 that was made, that the team executed as they 20 were asked to. 21 Q. In other inquiries or other significant 22 litigation that you're involved in, do those 23 conversations happen from the KPMG side? So do 24 you advise clients as to the pros and cons of 25 the different courses of action? 157 1 A. In some respects we do. With de-duplication, 2 which I think is the topic we're talking about, 3 you would almost always be involved at the 4 outset, which means that you could apply 5 de-duplication during the processing phase, 6 which has not been possible here. So we are in 7 a -- we're in a unique situation with the 8 dataset we have here. 9 Q. Given the uniqueness of this particular case, do 10 you think it would have been better for KPMG to 11 have advised on the pros and cons of that 12 particular course of action? 13 A. I think the clarification questions that we 14 asked were very clear and, yet again, I'm 15 comfortable that what we were asked to do, we 16 delivered on. Should we -- again, with 17 hindsight, would I have liked to have 18 highlighted some other implications or one major 19 implication of using this approach? Yes, 20 I would. 21 Q. When you say the clarification questions, can 22 you give us some examples of the kinds of 23 clarification questions that were posed? 24 A. Yes. So some searches would have been run. We 25 would have provided the results back to the 158 1 external solicitors. The external solicitors 2 would have said "These numbers still look 3 slightly too high", and then there would have 4 been a conversation, first about whether or not 5 the search terms were amended and, secondly, 6 about de-duplication. 7 When the conversation about de-duplication 8 happened, the clarification would be around what 9 would you like to be duplicated against -- 10 de-duplicated against what? And in this 11 particular case, because you had data sitting in 12 different cases, there was two types of 13 de-duplication of responsive items, one within 14 a dataset itself and then subsequently when you 15 move the data from various datasets to another 16 place, a second set of de-duplication in that 17 centralised location. But those conversations 18 were always about de-duplication of responsive 19 documents. 20 Q. I'd like to get a little bit of clarity in 21 respect of something we've heard this morning, 22 and that was about the effect that 23 de-duplication might have on further searches 24 and further requests? 25 A. Yeah. 159 1 Q. It seemed to be suggested that it affects the 2 body of available documents for other searches, 3 whether it be in respect of the particular 4 Rule 9 Request, let's say the de-duplication is 5 carry out in respect of question 11 but actually 6 we've got a question 12 that those documents 7 might be responsive to, or perhaps, in the 8 future, a further Rule 9 Request comes in and 9 the excluded document might be responsive to 10 that particular request but has been excluded as 11 a duplicate; can you tell us about that? Does 12 that actually happen? 13 A. No, it doesn't. So the de-duplication happens 14 only on a request or a question within a request 15 basis. It applies for that question within 16 a request and then no longer exists, 17 effectively, when the next question or the next 18 search is applied. 19 Q. Thank you. I want to move on to remediation. 20 Sorry, just before I do, though, in terms of the 21 numbers, you've said that often the response 22 from Herbert Smith Freehills would be to go for 23 a smaller number, to go for the one that has the 24 de-duplication applied, for example. 25 A. Yeah. 160 1 Q. In this particular Inquiry, have you experienced 2 a move towards reducing the number of documents 3 disclosed, as opposed to gathering the largest 4 number or a larger number? 5 A. Do you mean disclosed or -- 6 Q. Yes. As in are there pressures to disclose 7 smaller numbers rather than larger numbers? 8 A. No, not at all. I think the solicitors are 9 conscious of not overwhelming the Inquiry with 10 irrelevant data. That's the only reason they 11 would be lower. But absolutely not how you've 12 presented it. 13 Q. On remediation, you've obviously looked at the 14 problems that have occurred in this particular 15 case. What conclusions have you drawn, as KPMG, 16 as to the failings within the Post Office or 17 within Herbert Smith Freehills or somewhere 18 else, in respect of disclosure which ultimately 19 led to the issues that we are considering and 20 have been considering? 21 A. As I set out in one of the paragraphs in my 22 witness statement, it's the search terms and the 23 review of family documents which is ultimately 24 the reason why Appendix 6 didn't get picked up 25 back in April 2022. However, I clearly 161 1 recognised that the form of de-duplication is 2 also important and, if a review is not based on 3 responsive documents only, then it has 4 shortcomings. 5 Q. Have you noticed any change in respect of the 6 involvement of the Post Office in underlying 7 decision making? 8 A. Well, absolutely, with my interactions with 9 them, I have daily calls related to the 10 remediation and at least weekly or biweekly 11 calls related to the ongoing non-remediation 12 based activity. 13 Q. In respect of, for example, search terms and 14 future requests, have you seen any change in the 15 approach from the Post Office? 16 A. Well, I don't have interaction -- I don't have 17 visibility of the interaction between HSF and 18 the Post Office, but I have been on calls 19 regularly where it's been mentioned that those 20 have been shared for sign-off. 21 Q. Do you have confidence that the issues that we 22 have seen to date will not happen again and, if 23 so, why? 24 A. De-duplication, absolutely. We are only doing 25 family-level de-duplication now and we won't be 162 1 going back to item-level de-duplication, unless 2 there's a particular reason why that's 3 appropriate. I believe the review of family 4 items is ongoing and the workflow has changed 5 accordingly or the review workflow has changed 6 accordingly. Search terms I cannot give you 7 a guarantee on that, because the problem with 8 search terms is that they are not always precise 9 and it's difficult to know what you're looking 10 for until you find it sometimes. But I know 11 there's a lot of work being done to reflect on 12 the terms that have been used for previous 13 requests and whether they appeared appropriate 14 and fulsome enough. 15 MR BLAKE: Thank you. I don't have any further 16 questions. 17 Sir, do you have any questions? 18 SIR WYN WILLIAMS: No, I don't, thank you very much. 19 Thank you for your witness statement and for 20 coming to give evidence. 21 THE WITNESS: Thank you, sir. 22 MR BLAKE: Thank you very much, sir. Can we take 23 a break for 15 minutes now, so to return at 24 3.10, please. 25 SIR WYN WILLIAMS: Yes, certainly. 163 1 MR BLAKE: Thank you very much. 2 (2.55 pm) 3 (A short break) 4 (3.10 pm) 5 MR BEER: Good afternoon, sir, can you see and hear 6 me? 7 SIR WYN WILLIAMS: Yes, I can thank you. 8 MR BEER: May I call Gregg Rowan. 9 SIR WYN WILLIAMS: Yes. 10 GREGG NICHOLAS ROWAN (sworn) 11 Questioned by MR BEER 12 MR BEER: Mr Rowan, as you know my name is Jason 13 Beer and I ask questions on behalf of the 14 Inquiry. Can you give is your full name please? 15 A. Gregg Nicholas Rowan. 16 Q. Thank you for coming to the Inquiry today to 17 help us in our work and for previously providing 18 a witness statement and a long annex. You 19 should have in front of you a copy of that 20 witness statement, dated 23 August 2023. If you 21 look at page 63, you should see your signature? 22 A. I do. 23 Q. Are the contents of that statement true to the 24 best of your knowledge and belief? 25 A. Yes, they are. 164 1 Q. For the purpose of the transcript, the URN is 2 WITN09950100, there's no need to display that. 3 I am going to ask you about a range of matters 4 referred to in your witness statement but not 5 all of them. The Chairman has read your witness 6 statement and it will be introduced to the 7 public through disclosure on the Inquiry's 8 website. I am going to ask questions on two 9 principal issues: firstly, to understand more 10 about how and why discrete disclosure failings 11 occurred, concerning search terms, reviewing 12 families of documents and the de-duplication of 13 documents, so that's a backward look; then, 14 secondly, to seek to understand a little more 15 about what the Post Office and its advisers have 16 done to put things right for the future. 17 Can I start with your career qualifications 18 and expertise. You're a solicitor; is that 19 right? 20 A. A barrister. 21 Q. You're a barrister. You joined HSF in 2007; is 22 that right? 23 A. Yes, that's right. 24 Q. Becoming a partner in 2014? 25 A. Yes. 165 1 Q. I think you first began work on the Inquiry on 2 behalf of the Post Office in October 2021? 3 A. Yes, that's right. 4 Q. You took over from Andrew Lidbetter as the 5 designated or recognised legal representative 6 for the Post Office in April 2022; is that 7 right? 8 A. Yes. 9 Q. You are, I think, supported by a substantial 10 team within HSF, presently -- is this right -- 11 three partners? 12 A. Three more partners in addition to myself, yes. 13 Q. Yes. You're supported by project management 14 specialists, legal analysts, paralegals, 15 trainees, lawyers, of whom there are presently 16 150; is that right? 17 A. Yes, I think the number is slightly larger than 18 that, I think it's possibly increased towards 19 160, since I gave my statement. 20 Q. The legal team also consists of eight counsel, 21 including two silks; is that right? 22 A. Yes, that's right. 23 Q. Can we look at other advisers please, firstly 24 Peters & Peters. Can you summarise for us, 25 please, the role of Peters & Peters in relation 166 1 to the Inquiry? 2 A. Yes, Peters & Peters, having advised on all of 3 the criminal proceedings, appeals and related 4 matters, were instructed by Post Office to 5 advise in relation to the criminal aspects of 6 the Inquiry. Typically, that will involve -- 7 a substantial part of it involves responding to 8 Rule 9 Requests that relate to prosecutions but 9 there's a host of other work in addition to 10 that. 11 Q. So where does their role differ from that of 12 your own firm? 13 A. Until recently, I was the recognised legal 14 representative so my firm were the solicitors of 15 record in the Inquiry, and we would engage with 16 the Inquiry, receive the Inquiry's 17 correspondence, respond to its correspondence, 18 receive Rule 9 requests and respond to them. 19 Peters & Peters, I suppose their role is 20 a little bit more behind the scenes but doing 21 very similar work to the work we're doing, 22 albeit in the context of private prosecutions. 23 Q. How many lawyers do they have working on the 24 Inquiry. 25 A. I don't know, it's a relatively small team 167 1 supported by a larger junior barrister team. 2 Q. How large is the junior barrister team? 3 A. I don't know. 4 Q. Are we talking three, four, five or 10, 20, 30? 5 A. I think it's towards the lower end of that scale 6 rather than the -- 7 Q. Two, three, four, five? 8 A. That's right. 9 Q. Can we look, please, at something that Peters & 10 Peters undertook. It's on page 9 of your 11 witness statement and it's paragraphs 29 to 31. 12 I just want to spend some time narrating this, 13 almost, because it has relevance to later 14 events. I think that's right, isn't it? 15 A. Yes, it does. 16 Q. You tell us that: 17 "In late 2021, [Peters & Peters] were 18 instructed by [the Post Office] to undertake 19 preparatory work relating to issues 109 to 161 20 in the Inquiry's Completed List of Issues ..." 21 "CLI", as you call them. Just summarising 22 109 to 161 in our list of issues, all concern 23 the conduct of private prosecutions by the Post 24 Office. 25 A. Yes. 168 1 Q. Carrying on with the statement: 2 "This work comprised the identification of 3 material already available as part of the PCDE 4 ..." 5 Can you just explained what the PCDE is, 6 please? 7 A. That stands for the Post-Conviction Disclosure 8 Exercise, that's a substantial piece of work in 9 identifying documents relevant to the criminal 10 appeals, that includes case-specific documents 11 but also more general documents that have the 12 potential to cast doubt on a conviction. 13 Q. The PCDE was itself an exercise undertaken by 14 Peters & Peters? 15 A. Yes. 16 Q. So it continued as part of the PCDE that was 17 relevant to those issues: 18 "... preparing narrative documents that 19 provided information relevant to the issues; 20 obtaining additional documents and information; 21 and conducting the further application of search 22 terms and/or review of material related to 23 expects of the PCDE as necessary to collate 24 materials relevant to the issues. [Peters & 25 Peters] began this work in January 2022." 169 1 Paragraph 30: 2 "When [the Post Office] received [Rule 9s] 3 from the Inquiry relating to private 4 prosecutions, my firm worked with [Peters & 5 Peters] to prepare [the Post Office's] responses 6 to the requests. The level of involvement of 7 [Peters & Peters] has varied, depending in 8 particular on the information and/or documents 9 sought by the Rule 9 Requests. For certain 10 Rule 9 Requests, [the Post Office] and my firm 11 have utilised [Peters & Peters'] prior work 12 product." 13 Then 31: 14 "For the Rule 9 Requests in which [Peters & 15 Peters] have been involved, members of my team 16 have worked closely with members of theirs, both 17 in terms of agreeing the general approach on how 18 to respond to the requests and in advising [the 19 Post Office] as our mutual client. Until very 20 recently, my firm has not generally been 21 involved with [Peters & Peters'] conduct of 22 reviews, including their devising of search 23 terms, their consideration of whether or not to 24 review family documents, or their instructions 25 to KPMG ... in respect of de-duplication when 170 1 running searches." 2 So just stopping there, I'm not going to 3 deal with the exception that you then don't deal 4 with. 5 Just stopping there, are you saying there 6 that, until recently, by which I take to mean 7 after discovery of the disclosure issues that 8 we're looking at today, your firm did not have 9 visibility on what Peters & Peters had done in 10 terms of the settling of search terms, the 11 reviewing of family documents or their 12 interactions with KPMG on de-duplication. 13 A. Generally not. It's possible that we had sight 14 of some of their search terms. 15 Q. Sight of them meaning incidentally they might 16 have been disclosed in an email or a letter, as 17 opposed to "We're proposing to apply these 18 search terms, do you think they're good, have 19 you improvements to make"? 20 A. Yes, that's right. 21 Q. That kind of visibility? 22 A. Yes. 23 Q. What has changed, therefore, recently in 24 relation to those three issues? 25 A. Well, principally, what we've been doing 171 1 recently is the remediation and the assurance 2 work and as part of aspects of that, we have 3 worked with Peters & Peters to devise search 4 terms. Broadly speaking, we've had more active 5 involvement. 6 Q. What about family documents in relation to work 7 that they are doing? 8 A. Well, I think because the approach to family 9 documents is largely now settled, there's 10 an established approach, there can be variation 11 in it but, generally, the approach is to review 12 family documents. I think that's common ground 13 and understood. 14 Q. I'm winding forward to another topic but whilst 15 you've mentioned it, you said that the approach 16 to family documents is now firmly established. 17 When we heard from Mrs Wills this morning, she 18 said that it was -- for a backward look, it was 19 established that family documents were now to be 20 reviewed where they hadn't been reviewed in 21 previous Rule 9s but, in the future, an approach 22 hadn't been settled and it was going to be done 23 on a case-by-case basis. Is that correct or 24 incorrect? 25 A. I think it's broadly correct. Generally 172 1 speaking, the approach will be to review 2 families but, within that proposition, there are 3 different ways of going about it and each Rule 9 4 Request will have to be considered on its merits 5 to decide precisely how to approach it. But, 6 broadly speaking, yes, the approach going 7 forward will be to review families, and, if 8 I may, I would clarify that may be the families 9 only of keyword responsive documents that are 10 identified as being relevant or it may be all 11 families, depending on the circumstances. So 12 they're two of the main alternatives that might 13 be pursued. 14 Q. We'll come back to that a little later. You say 15 here that your firm was not, until recently, 16 aware of the instructions given to KPMG by 17 Peters & Peters about de-duplication. Did you 18 know that Peters & Peters were giving 19 instructions to KPMG about de-duplication? 20 A. Personally, I wouldn't have -- I wouldn't have 21 known that. I think, whether the HSF team were 22 aware of that specifically, I don't know. 23 I think it could be inferred that they would be 24 giving instructions about de-duplication because 25 that's precisely the sort of thing that, in 173 1 an eDisclosure exercise, would be part of the 2 instructions. 3 Q. Surely, you would want to ensure a consistent 4 approach between the firms? 5 A. I think it's difficult to speak in such general 6 terms. Each Rule 9 Request and often questions 7 within Rule 9 requests have to be dealt with on 8 their own merits and it's a multifactorial 9 assessment where there are so many different 10 inputs. We certainly shared a common objective 11 of fulfilling Post Office's disclosure 12 obligations but how to do that will vary so much 13 between requests that I think it's difficult to 14 speak of a common approach. 15 Q. Wouldn't you at least wish to know what each 16 firm was doing in terms of de-duplication? 17 A. Not necessarily, no, I don't think so. 18 Q. Well, in the absence of knowledge -- or is this 19 absence of knowledge of what Peters & Peters 20 were instructing KPMG to do, one of the reasons 21 why the de-duplication of errors has taken 22 a significant time to put right, because you 23 have had needed to find out from KPMG exactly 24 what had happened in the first place? 25 A. Yes, between ourselves and Peters & Peters, we 174 1 have had to understand from KPMG this -- the 2 occasions on which they applied item-level 3 de-duplication. I think that would always have 4 been necessary. 5 Q. Has that been a very simple and straightforward 6 task in understanding from KPMG when they 7 de-duplicated and, if they did, which form of 8 de-duplication they applied? 9 A. No, it hasn't. 10 Q. Has it been the precise opposite of that? 11 A. Yes, it has. 12 Q. Can you explain why that is? Why it has been 13 problematic and difficult? 14 A. My understanding is that there hasn't been 15 a clear auditable record of the occasions on 16 which item-level de-duplication was applied, so 17 it's had to be ascertained from a very forensic 18 and detailed review of contemporaneous 19 materials, and that has taken a long period of 20 time. 21 Q. We just heard from Mr Tombleson who sat in the 22 same chair as you saying that there was a clear 23 record disclosed by -- because of the unusual 24 nature of the request for item-level 25 de-duplication, that there was a clear record in 175 1 exchanges between associates and senior 2 associates and his managers and senior managers. 3 From your perspective, is that correct? 4 A. Certainly when you find the emails you can see 5 what has been instructed and you can see the 6 dialogue between our senior associates and 7 KPMG's managers. So, to that extent, there is 8 a record, yes. 9 Q. Given that there is a record, why has it been 10 difficult and problematic to work out when 11 de-duplication has occurred and, if so, what 12 form that de-duplication has taken? 13 A. My understanding is that simply to find the 14 emails in -- between ourselves and KPMG, there 15 are probably hundreds, possibly many hundreds, 16 of emails on a daily basis. It's a constant 17 dialogue. And my understanding is that for KPMG 18 to identify the relevant instructions from those 19 emails has taken time. I should say, I'm 20 explaining my understanding. Obviously KPMG 21 would be -- they would be able to explain 22 precisely what steps they've taken. 23 Q. I think we just heard from Mr Tombleson, who 24 said actually it was quite straightforward. Has 25 that been your experience? 176 1 A. No, it hasn't. 2 Q. Can we turn to KPMG, please, on page 8 of your 3 witness statement at paragraph 25. You say that 4 the Post Office's eDiscovery provider is KPMG, 5 who are directly accountable to the Post Office 6 for their work. Does that first sentence mean 7 that KPMG liaise directly with the Post Office 8 about eDisclosure directly, because they were 9 directly accountable to them? 10 A. At a strategic level in relation to cost, in 11 relation to resourcing, as others have mentioned 12 today, that dialogue would take place 13 principally between KPMG and Post Office. 14 Forgive me, in terms of instructions on 15 specific items of eDiscovery workflow, those 16 instructions would come from ourselves or from 17 Peters & Peters. 18 Q. So KPMG worked according to instructions given 19 to them by HSF on a Rule 9 by Rule 9 basis? 20 A. Yes, and Peters & Peters, yes. 21 Q. If KPMG were working to your instructions and 22 those of Peters & Peters, rather than from 23 instructions given on a detailed basis by the 24 Post Office, how did the Post Office maintain 25 oversight and supervision of what was being 177 1 carried out in its name? 2 A. I can only speak to our engagement with Post 3 Office in relation to eDiscovery and KPMG's role 4 and eDiscovery issues were a subject that we 5 discussed intermittently when there were points 6 to discuss. 7 Q. To your knowledge, was there any intrusive 8 oversight and supervision directly from the Post 9 Office of the way that KPMG was carrying out its 10 work? 11 A. No, there wasn't. I should say nor would 12 I expect there -- to see that sort of 13 examination of what an eDiscovery provider is 14 doing. 15 Q. To your knowledge, were there any communications 16 before these issues were discovered from either 17 Peters & Peters or HSF back to the Post Office, 18 explaining the way in which de-duplication was 19 being carried out? 20 A. No. 21 Q. Same question in relation to whether or not 22 families of documents were being reviewed? 23 A. Generally speaking, the approach of my firm -- 24 I can't speak for Peters & Peters, but for my 25 firm -- was to send regular email updates to 178 1 Post Office explaining in detail the steps that 2 we were taking in the conduct of their 3 disclosure and, in those emails, we identified 4 that we were reviewing keyword responsive 5 documents. 6 Q. I missed a word there, you were reviewing? 7 A. Keyword responsive documents, hits. 8 Q. Thank you. 9 A. So to that extent, we explained the position to 10 Post Office. 11 Q. But nothing telling them of the nature of the 12 de-duplication that had happened before your 13 review nor of -- 14 A. No. 15 Q. -- the extent to which a reviewer might search 16 for or not search for family documents, as part 17 of their review? 18 A. Well, certainly not in relation to 19 de-duplication and I think I've answered the 20 question in relation to family documents. What 21 we explained was that we were reviewing keyword 22 responsive documents. 23 Q. So you didn't tell the Post Office that you were 24 not reviewing family documents? 25 A. Yes, yes. 179 1 Q. Was there any reason for that, for not telling 2 them this? 3 A. I don't think there was a particular reason. 4 I think that the explanation that we were 5 reviewing keyword responsive documents, that 6 indicated the approach that we were taking. 7 Q. But you could say to a client, "Look, there's 8 two ways of conducting a de-duplication 9 exercise, or at least two ways of conducting 10 a de-duplication exercise. We've got a very 11 difficult dataset here because of the sources of 12 it, that adds a layer of complexity. We could 13 do or instruct KPMG to do item-level 14 de-duplication, that's has these consequences, 15 or we could instruct them to do family level 16 de-duplication, that has these consequences -- 17 a larger pool of documents" -- 18 A. Yes. 19 Q. -- "that will have these cost consequences". 20 From what I'm understanding, none of that 21 happened in relation to de-duplication? 22 A. No, and, in fact, I don't think members of my 23 team had in mind that there were different types 24 of de-duplication that might be applied. 25 Q. We'll come to examine that in a moment. In 180 1 relation to the review of family documents, 2 again, have I understood you correctly to say 3 that there wasn't communication back to the 4 client about when and in what circumstances that 5 was occurring? 6 A. Yes, that's right. 7 Q. What was the reason for that? 8 A. I think that the way the updates were drafted 9 was simply -- the updates generally were 10 incredibly detailed and I think the way that the 11 approach was expressed was that we're reviewing 12 keyword responsive documents and it was a form 13 of drafting more than anything else. 14 Q. It was a form of? 15 A. Drafting. It was the way in which the 16 explanations were drafted. 17 Q. That only tells half a story, doesn't it? It 18 says what you are doing but it assumes that the 19 client knows what you're not doing. 20 A. I think, with the benefit of hindsight, I think 21 that's fair. 22 Q. Can we move forward to the stage when you get to 23 the review, please, so after a de-duplication 24 exercise has taken place, and look at 25 paragraph 48(e) on page 18 of your witness 181 1 statement. 2 If we scroll down please, and I'm picking 3 this up partway through the exercise and you're 4 telling us here about after a de-duplication 5 exercise has happened, irrespective of the 6 species of de-duplication, when the pool of 7 documents is given back to your firm, what then 8 happens; is that right? 9 A. Yes. 10 Q. You say: 11 "The first level review is typically 12 conducted by reviewers in our ALT (Alternative 13 Legal Services) team or an associate team in 14 London and/or elsewhere in my firm's global 15 network." 16 So this is the first level of review. The 17 Alternative Legal Services team, where are they 18 located? 19 A. They have hubs around the world, principally the 20 team working on this matter has been based in 21 Belfast and Johannesburg. More recently, we've 22 involved resource in Melbourne, Australia. 23 Q. So Belfast, Johannesburg and more recently 24 Australia? 25 A. Yes. 182 1 Q. Are the members of the Alternative Legal 2 Services team conducting the reviews for the 3 Post Office lawyers? 4 A. Sorry, I don't entirely understand the question. 5 Q. Are they lawyers? 6 A. The first level reviewers are typically made up 7 of what we call legal analysts and senior legal 8 analysts. They tend to be law graduates, 9 they're supervised by lawyers. 10 Q. So law graduates but not qualified in their 11 jurisdiction or our own? 12 A. Yes. 13 Q. So they don't hold a practising certificate? 14 A. No. 15 Q. Are they dedicated to working on this Inquiry or 16 might they be working on a number of disclosure 17 exercises at once? 18 A. Generally, we had a core team that was committed 19 to working on the Inquiry. There were occasions 20 where there were lulls, in which case members of 21 the team would work on other things. There were 22 occasions where, such as the present one, where 23 we needed to drawn in more resource, in which 24 case we bring in people from -- that may have 25 been working on different things. 183 1 Q. In a moment we're going to come to look at the 2 reviews of families of documents and, in 3 particular, whether these reviewers were 4 instructed to look at documents that were within 5 a family of which a responsive document was 6 a part. Can we just jump ahead to that issue 7 now and see what you say about it, in 8 paragraph 95 on page 37 of your witness 9 statement. You say: 10 "It should be noted that when a decision is 11 taken that family documents need not be reviewed 12 that does not mean that they cannot be 13 considered. Unless they have been removed in 14 the de-duplication process, the entire family of 15 each document to be reviewed is still available 16 to our reviewers whilst not reviewing family 17 members will be the default when that decision 18 has been made. If the reviewer considers the 19 content of a document within a family is such 20 that there might be other documents within that 21 family which should be reviewed to understand 22 the context of responsive document, then they 23 will proceed to conduct that further review." 24 The reviewer that's being referred to here, 25 is that right, the individual within, in the 184 1 first instance, the Alternative Legal Services 2 team? 3 A. In the first instance, yes. 4 Q. So you've got these people in other parts of the 5 world who were graduates in law, but not legally 6 qualified, making decisions on whether to look 7 at family documents or not? 8 A. For the purpose of assessing whether the 9 document they're actually looking at is 10 responsive or possibly privileged. 11 Q. Not for the purpose of establishing whether 12 there are other documents within the family that 13 might be relevant to the request that they are 14 addressing? 15 A. Generally, that wouldn't be their role as first 16 level reviewers no. 17 Q. Why wouldn't it be their role? 18 A. The way that reviews are set up, generally 19 speaking they conduct linear reviews of 20 documents that are batched to them, documents 21 might be batched in, say, sets of 50, and the 22 expectation of the reviewers is that they review 23 the documents in front of them and they're not, 24 if you like, going on a train of enquiry. If 25 they see a particular document, it's not their 185 1 role to look at that document and to see where 2 it takes them, as it were. 3 Q. So although what you said in paragraph 95 is 4 correct, that the reviewing pane displayed to 5 the reviewer would allow them to go and look at 6 the family, in practice, that's not their 7 function? 8 A. Well, they would look at the families, actually. 9 They're encouraged to look at families in order 10 to determine the relevance of the document 11 they're actually looking at. It may be that 12 it's -- it may be that they want to understand 13 something more about the documents and they 14 think it is helpful to go and look at a family 15 member. So, in those circumstances, they're 16 encouraged to do that. Whether or not they do 17 it will depend on the judgement of the 18 individual reviewer in the given case. 19 Q. Were written instructions given to these people 20 within the ALT about when and in what 21 circumstances they should access the facility 22 within the reviewing pane to look at family 23 documents? 24 A. Part of their workflow, when we're conducting 25 a hits-only review, is to look to families when 186 1 they think that will be helpful to determine 2 whether a document is relevant or privileged for 3 context, possibly. That is something that they 4 routinely do when conducting hits-only reviews. 5 I have to say, I don't know whether that 6 instruction is in writing or given verbally but, 7 from the enquiries that I've made, I'm satisfied 8 that it's given. 9 Q. But they're not the people we should be 10 expecting to go on a train of enquiry? 11 A. No. 12 Q. Who are the people we should be expected to go 13 on the train of enquiry? 14 A. If anybody goes on a train of enquiry, it would 15 be at the second-level review stage. So, 16 typically, the documents that are identified at 17 first-level review as being relevant are then 18 reviewed by lawyers, principally based in 19 London, almost inclusively based in London. 20 Q. But, by that stage, a lot of documents will have 21 been excluded including their families by the 22 first tier of reviewers, won't they? 23 A. Yes, that's right. 24 Q. So the second tier of reviewers don't have 25 a chance of looking at those families, have 187 1 they, because they don't know that they exist, 2 they're not within their batch? 3 A. Well, there is some QC over documents that have 4 been tagged as not being relevant. So people do 5 look at those documents but for QC purposes. 6 Q. But is the QC process, the quality control 7 process, picking up whether the first-level 8 reviewers are appropriately following a train of 9 enquiry? 10 A. (The witness shook his head) 11 Q. No, because that's what their job? 12 A. No, precisely. 13 Q. So the QCing process is not going to bring more 14 documents back within the pool on the basis of 15 a failure to look at families? 16 A. No, that's right. 17 Q. So if the second and perhaps third-level 18 reviews -- which is lawyers in London, 19 essentially? 20 A. Yes. 21 Q. Do they include, amongst their tasks, reviewing 22 whether the ALT staff abroad have appropriately 23 looked at family documents? They can't, can 24 they? 25 A. When you say "appropriately looked at family 188 1 documents"? 2 Q. Yes, whether they have found a document that's 3 responsive to a search term, a hit, looked at 4 the reviewing panel, seen that there are 5 25 other documents of which that's a family, 6 gone into those other documents, seen that 7 they're actually relevant to this Rule 9 8 Request, "We'd better disclose those"? 9 A. No, they wouldn't be looking -- they wouldn't be 10 conducting their analysis of the document for 11 that purpose. 12 Q. So who does, then? Who checks families for 13 relevant material? 14 A. Well, it is open to the second-level reviewers 15 to turn to review the families of keyword 16 responsive documents that come through to 17 second-level review. So if they're to be 18 reviewed, it would be at that stage. 19 Q. But they're looking at an already narrowed pool? 20 A. Yes, they are. 21 Q. Is that still the position today? 22 A. The -- there are different approaches to 23 reviewing family documents. Either -- the two 24 main ones -- and I mentioned this earlier, are 25 two -- when conducting a review, to look at all 189 1 of the keyword responsive documents and all of 2 their families. The alternative is to look at 3 the keyword responsive documents that are tagged 4 as being relevant and their families. Both of 5 those options are being used. The teams have to 6 judge in individual circumstances which of those 7 options to pursue. 8 Q. What determines that judgement? What are the 9 relevant factors? 10 A. There are a host of factors. Obviously a key 11 one is the expectation based on the knowledge 12 that the senior associate setting up the review 13 have of the dataset and the issues, but the 14 expectation is to the likelihood of identifying 15 relevant documents in the family document set. 16 And, again, that will vary case by case. It 17 will also depend very significantly on the 18 number of family documents that exist. 19 In this matter, what we call the family 20 multiplier has been very high, that's the number 21 of additional family documents you get in 22 addition to the keyword responsive documents. 23 Q. The family multiplier question, how is that 24 answered? Is it "If the family multiplier is X 25 or above, we won't do it, whereas if it's Y or 190 1 below, we will"? 2 A. No, it's really not as simple as that. There 3 are so many different factors that inform 4 a decision as to how to approach a review and 5 it's a very dynamic situation, decisions have to 6 be made iteratively and sometimes revisited. 7 There's no hard and fast rule and there really 8 shouldn't be. It's a much more thoughtful and 9 careful exercise. 10 Q. Can we return to keyword searches before again 11 going back to the family issue because the two 12 feed into each other. If we turn up 13 paragraph 48(a) of your witness statement which 14 is on page 17. If we just scroll down, you set 15 out the process. At the beginning of the 16 process, you say: 17 "We send each request [Rule 9 Request] to 18 [the Post Office] and discuss with [the Post 19 Office] what is likely to be required, including 20 which of the relevant methods above are likely 21 to yield relevant documents, identifying the 22 repositories that need to be searched ... the 23 extent to which harvests of potentially relevant 24 documents have already been undertaken [and then 25 this] formulating keyword [searches], and 191 1 harvesting and processing and uploading to 2 Relativity of any additional materials." 3 That suggests that currently you discuss 4 with the Post Office the formulation of keyword 5 searches, correct? 6 A. Yes. 7 Q. Would you agree that that's the sensible and 8 indeed necessary approach, because the Post 9 Office and its employees are most likely to be 10 in possession of relevant knowledge to be able 11 to develop accurate and reliable search terms? 12 A. No, I don't think I would agree with that. 13 Often, when we devise keyword searches, it's 14 based on prior discussions with subject matter 15 experts within the Post Office business. So if 16 I could give you an example, in the context of 17 the civil proceedings aspect of this phase, one 18 of the early things that members of my team did 19 was to go out and speak to a large number of 20 people, subject matter experts in the Post 21 Office business, I think there were about 14 of 22 them, and obviously that's a multifaceted 23 conversation, talking about documents, trying to 24 understand the issues. 25 It's those sorts of conversations that are 192 1 the basis for members of my team and I should 2 say there are many other things as well, 3 familiarity with documents being another obvious 4 one. There are many different inputs but it's 5 on the basis of those inputs they would then 6 devise search terms and then sent those search 7 terms to the Post Office. 8 Q. Has that always been the case: that you've 9 involved, to the maximum extent possible, as 10 a firm, the Post Office in the development of 11 search terms? 12 A. Look, I wouldn't like to say to the maximum 13 extent possible. That's -- I think that 14 probably puts it too high. But certainly what 15 we have always sought to do is engage 16 extensively with people within the Post Office 17 business to understand the issues that are the 18 subject matter of the Inquiry, issues on which 19 they have background or expertise, and they will 20 inform the approach that we take. 21 We would also obviously be in dialogue with 22 members of the Post Office Inquiry Team. 23 Q. So it's the case that in the Rule 9 Requests 24 that have been served on the Post Office by this 25 Inquiry, the Post Office has been involved 193 1 appropriately at all stages in the development 2 of search terms to respond to them, and it's not 3 a matter where the Post Office can say, "That's 4 down to our legal advisers and not us"? 5 A. I think there's a lot in that question. 6 Certainly, we have, necessarily in our role, had 7 extensive engagement with the Post Office 8 business and, again, as I've said, we've sought 9 to understand documents issues, and a part of 10 that has been informing the content of search 11 terms. I have to say, I wouldn't necessarily 12 expect Post Office as a client to be commenting 13 in detail on the search terms that we devise. 14 Some clients might but others might not. So 15 I don't think it necessarily follows that 16 because Post Office weren't actively commenting 17 on search terms, that -- or that a client isn't 18 actually commenting on search terms, that that 19 should be seen as being inappropriate. 20 Q. Can we look at paragraph 61, please on page 23 21 of your witness statement. It's the last 22 sentence, you say: 23 "The search terms are generally shared with 24 [the Post Office], although I do not believe 25 that we received specific input or feedback from 194 1 the Post Office in relation to the search terms 2 now under discussion." 3 When you say "The search terms are generally 4 shared with [Post Office]" does that mean that 5 on some occasions your firm does share search 6 terms with the Post Office and other terms (sic) 7 keeps them to themselves and doesn't share them. 8 A. I would say that we share them in all or nearly 9 all circumstances. I can't say absolutely that 10 they have always been shared. 11 Q. So generally should be read as all or nearly 12 all? 13 A. Yes. 14 Q. Okay. What determines whether search terms are 15 not shared or would it just be oversight if they 16 haven't? 17 A. I think I would be, yes. The intention would be 18 to share search terms. 19 Q. You say: 20 "... I do not believer that we received 21 specific input of feedback from POL in relation 22 to the search terms now under discussion." 23 Are you relating that sentence, ie "under 24 discussion", to Rule 9(11) and (14) and the 25 relevant part of them? This comes within 195 1 a section of your witness statement, I should 2 say, which is talking about the use of search 3 terms generally, not when you turn later to look 4 at Rule 9(11) and 14? 5 A. Yes. I think I am and I think I probably 6 have -- I have in mind, in particular, the 7 search terms relating to request 11, questions 8 15 and 46(a) which were the ones that my firm 9 devised. 10 Q. So that sentence there "now under discussion" 11 means "insofar as I'm discussing the relevant 12 parts of Rule 9(11) and (14)"? 13 A. Yes. 14 Q. Okay. You say that you don't believe that there 15 was any specific input or feedback from POL in 16 relation to those. How frequently did the Post 17 Office give input or feedback in relation to the 18 search terms that were shared with it? 19 A. I don't recall whether that happened at all. 20 Q. So for the last couple of years, despite sharing 21 search terms with the Post Office on all or 22 nearly all occasions, say for oversight, the 23 Post Office has never come back and said, "Hold 24 on, what about X or Y? I don't think that's 25 appropriate. You might want to look using this 196 1 word"? 2 A. No. Save, I should clarify, that recently, as 3 these issues have come to light, and we've been 4 focusing on remediation, the Post Office team, 5 in the way that Mrs Wills described, has been 6 more active and engaged in relation to search 7 terms. 8 Q. Is your impression that's because of the issues 9 that have been uncovered or maybe even a change 10 of personnel? 11 A. I think it's -- the issues that have been 12 uncovered is certainly a factor. 13 Q. So a more proactive and involved client, would 14 that be a fair way of describing it? 15 A. Very much so. 16 Q. Turning from the general, then, to the specific 17 and the use of the search terms in relation to 18 parts of Rule 9(11) and (14). I wonder whether 19 we can turn to paragraph 65 on page 24, which is 20 over the page. It's the foot of the page. You 21 say: 22 "The particular sub-questions relating to 23 prosecutions or investigations policies were 15 24 and 46(a) ..." 25 This is Rule 9(11). Then over the page to 197 1 page 25, please, you set them out and if I just 2 read the relevant parts. 15, last sentence: 3 "Please provide copies of the same [that's 4 prosecutions policy] and copies of all 5 iterations of the prosecutions policy since 1999 6 that are in POL's custody or control." 7 Then 46(a): 8 "Please provide copies of the same [that's 9 an investigations policy] and copies of all 10 iterations of the investigations policy since 11 1999 that are in POL's custody or control." 12 You say that you used, or your firm used, 13 three principal methods to identify documents 14 that were responsive to those two requests. If 15 you go to the foot of the page, please. At 68, 16 method one, you say: 17 "First, it was clear that these documents 18 would be sought by the Inquiry, so in January 19 2022 a workstream (the 'Policy Review'), was 20 scoped by [Peters & Peters] together with [the 21 Post Office], and on which my team was asked to 22 comment, one aspect of which was targeting the 23 documents referred to in CLI [that's the 24 Concluded List of Issues] 109." 25 That is set out in paragraph 66 above. We 198 1 can see it on the page: 2 "This was part of the preparatory work 3 relating to issues 109 to 161 in the [Concluded 4 List of Issues] that [the Post Office] 5 instructed [Peters & Peters] to undertake, and 6 one of a number of workstreams that [the Post 7 Office] has instructed [Peters & Peters] to 8 carry out in connection with criminal matters 9 that are relevant to Phase 4. 10 "69. As part of [that, Peters & Peters] 11 utilised the work that they had undertaken as 12 part of the PCDE ... I understand that the 13 approach to the Policy Review was adopted in 14 anticipation of [the Post Office] receiving 15 Rule 9 Requests ... When [the first request was 16 issued Peters & Peters] were in the early stages 17 of the Policy Review, and we liaised with them 18 ... in order to draw upon their work when 19 responding to Questions 15 and 46(a)." 20 So essentially saying the first method was 21 that work had already been undertaken or was 22 being undertaken by Peters & Peters by reference 23 to the concluded list of issues, paragraph 109. 24 Had that turned up Appendix 3? 25 A. Yes, I think it had. My recollection is that it 199 1 had done, and we actually wrote to the Inquiry 2 setting out a list of documents that we'd 3 identified that we didn't think were responsive 4 to the particular requests. The requests were 5 focusing on very specific documents, the 6 prosecutions policy and the investigations 7 policy. We knew what those documents were, we 8 had various iterations, so did Peters & Peters. 9 So we knew what we were looking for. 10 But in the process of the exercise that 11 I describe in these paragraphs of my statement, 12 miscellaneous other documents had been 13 identified, and I think I'm right in saying it 14 that I think it was Peters & Peters who 15 identified Appendix 3 and, again, we wrote to 16 the Inquiry drawing -- 17 Q. Was that part of a suite of documents, a family 18 of documents, when it was identified? 19 A. If it was in Peters & Peters -- forgive me, 20 I don't recall which -- there are a number of 21 iterations of Appendix 3 and I don't recall, as 22 I sit here, which one specifically was 23 identified. What we know, though, is that the 24 iteration of Appendix 3 that formed part of the 25 PCDE was a standalone version, so it didn't 200 1 have -- 2 Q. It didn't have Appendix 6 as part of its family? 3 A. No, it didn't have any family members. 4 Q. So do you think that's the version of Appendix 3 5 you're talking about now? 6 A. I think it is but if I misremembered then I'll 7 correct that subsequently. 8 Q. You say in paragraph 70, if we scroll down, the 9 second method was liaising with the Post Office 10 to try to find relevant documents. This 11 included consideration of a FOIA request in 2020 12 that the Post Office had received seeking the 13 "current guidance and/or rules provided to POL 14 prosecutors and investigators": 15 "As part of our work ... we liaised 16 extensively with POL and others to try to 17 understand the policies that were in place at 18 various times. We had been in contact with 19 POL's Head of Security Operations and we also 20 sought information directly from [Peters & 21 Peters] and Cartwright King ... Members of [the 22 Post Office's FOIA] request team also spoke 23 directly with ... individuals in POL's business 24 including the team lead investigator, the 25 manager of the intelligence and administration 201 1 team, and the individual who subsequently 2 identified the suite of appendices concerning 3 [sic] Appendix 6 ... At that time, we were 4 provided with various iterations of the 5 prosecutions policy and versions of a 'conduct 6 of criminal investigations' policy." 7 By that, are you saying that those 8 approaches -- or did those approaches turn over 9 a version of Appendix 3? 10 A. No, I think the work that I'm describing there, 11 responding to the Freedom of Information Act 12 request that was concerned with the 13 prosecutions, I think it was concerned with the 14 prosecutions and investigations policy and 15 I don't think it had turned up Appendix 3. 16 Q. Nobody that was approached said "Hold on, there 17 is this collection of documents which we used 18 when we prosecuted anyone. They're part of 19 a collection of documents, and they're all in 20 a zip file that was commonly distributed to all 21 and sundry and, by the way, there's this 22 racially offensive document within them"? 23 A. I don't think that happened in the FOIA context. 24 I think in the policy review that I describe at 25 paragraph 69, as part of that exercise, Peters & 202 1 Peters spoke with various people within Post 2 Office's Security team. I understand that in 3 one of those conversations, the suite of 4 documents was referred to in general terms, no 5 reference to racist or offensive language, but 6 it was referred into general terms. 7 Q. Was that chased down or not? I assume not? 8 A. Well, it was established in the context of that 9 conversation that Peters & Peters already had 10 harvested the suite of documents, as they did. 11 What goes wrong is that the version of the 12 suite -- or the version of Appendix 3 in the 13 PCDE just so happens not to have any family 14 members. 15 Q. So did nobody put two and two together and say, 16 "Hold on, we've been told about a suite of 17 documents which was essentially a compliance 18 requirement, you had to do the things in the 19 suite if you wanted to prosecute a person. 20 We've got a bit of that, but we haven't got the 21 rest of it"? 22 A. My understanding, and, again, this is work that 23 Peters & Peters did, is that they'd understood 24 that the suite of documents -- and I don't think 25 it was described in those terms, but what we're 203 1 now calling the suite of documents -- existed. 2 They satisfied themselves in conversation with 3 the member of the Post Office Security Team that 4 they'd harvested it and I think it was believed 5 that, because of the breadth of the PCDE, 6 insofar as they'd harvested it and it was in any 7 way relevant, it would have been in there. 8 That's my understanding of the position. 9 Q. So the answer is that there was a belief that 10 they must have had it but, in fact, it wasn't 11 accessed or reviewed for the purposes of 12 responding to the Inquiry's request? 13 A. Yes, neither Peters & Peters nor my firm 14 reviewed Appendix 6. 15 Q. Then the third method, if we follow down at 16 paragraph 71, is the use of search terms, and 17 you set out what they are. We've looked at that 18 already so I'm going to skip over the page to 19 paragraph 72, please. You say: 20 "Looking back at the search terms and the 21 search criteria, whilst I appreciate that 22 Appendix 6 was not responsive to them, even with 23 the benefit of hindsight and knowing what I do 24 now about Appendix 6 and the suite of documents 25 to which it belonged, doing my best to be 204 1 objective, I consider these to be a reasonable 2 set of search terms ... The Inquiry's request 3 was for ... two specific documents ... and the 4 search terms sought to home in specifically on 5 prosecution or investigation policies. 6 Appendix 6 is in many ways an extraordinary 7 document. It contains no words about policy, 8 procedures or guidance. Using search terms 9 targeting policies and procedures, it would only 10 be possible to generate search terms to which 11 Appendix 6 responded if one knew about the 12 existence of the document first." 13 So that's essentially the same point that 14 Mrs Wills was making this morning, in a slightly 15 less direct way. But what you said there only 16 focuses on the reasonableness of the search 17 terms, doesn't it? 18 A. Yes. 19 Q. I think you would agree that search terms, or 20 viewing search terms as being the beginning and 21 the end of a search for relevant materials is 22 a folly? 23 A. Not necessarily, no, I wouldn't agree with that. 24 Q. You wouldn't? You think that can be appropriate 25 to use and only to use search terms and only 205 1 disclose the documents that are responsive to 2 the search terms? 3 A. Yes, I do. 4 Q. Why? 5 A. For example, if the request is for a specific 6 document, and many of the Inquiry's requests 7 were for specific documents, you devise a set of 8 search terms that you think will identify the 9 document and you run the search terms and you 10 find the document, then in those circumstances, 11 my view would be that the search terms are 12 appropriate or, taking the questions 15 and 13 46(a), the request was for all iterations of two 14 documents, the prosecutions policy and the 15 investigations policy. It seemed to us that, in 16 that context, as part of a broader set of steps 17 that were appropriate to take, in terms of 18 catching all iterations, search terms really had 19 an essential role. 20 We had the policy, we knew what it was 21 called, we knew the words that were contained in 22 it, and we were confident that using certain 23 search terms would find other iterations of it. 24 Q. So what about Request 14? If we go forwards, 25 please, to page 31 of your witness statement. 206 1 At the foot of the page, paragraph 83, 2 Request 14 materially asks at question 18 for: 3 "Policies and guidelines ... relating to the 4 bringing of private prosecutions against 5 subpostmasters and other end users ... alleged 6 to be responsible for shortfalls shown by 7 Horizon during the relevant period." 8 That's not seeking a specific document nor 9 iterations of that specific document, is it? 10 A. No -- 11 Q. It's more broadly crafted? 12 A. Yes, I agree. 13 Q. What did your firm do in response to the receipt 14 of paragraph 18 of Rule 9 Request 14? 15 A. In respect of the questions in Request 14 that 16 were focused on prosecutions, they were led by 17 Peters & Peters. So, on receipt of the request, 18 we shared it with Peters & Peters. As always, 19 when a request comes in, there's a frenzy of 20 activity and dialogue as to how to respond to 21 it. Peters & Peters then, I think by this 22 stage, the policy review exercise that 23 I described earlier in my statement, I think had 24 been completed or, if it hadn't been, it was in 25 significantly advanced form, and its conclusion 207 1 was that the PCDE would be sufficient for the 2 purpose of responding to issue -- I think it was 3 109 in the Inquiry's list of issues. 4 So my understanding is that Peters & Peters 5 then focused on the contents of the PCDE, 6 insofar as it relates to policies and 7 guidelines. 8 Q. So is the answer that your firm didn't do 9 anything in relation to this request, it was 10 passed to Peters & Peters to administer? 11 A. Yes. 12 Q. You don't know, I think, the search terms that 13 Peters & Peters were running. You didn't know? 14 A. I don't think we did at the time. 15 Q. Has a problem with those been identified? 16 A. Yes, it has. 17 Q. What's the nature of the problem that has now 18 been identified? 19 A. The -- ordinarily with a request such as this, 20 that seeks policies and guidelines, as a minimum 21 the search terms would contain a stem of the 22 word "policy" and a stem of the word 23 "guidelines". 24 Q. What I've described as the truncated version of 25 the word? 208 1 A. Yes, precisely, and the search terms that were 2 used in that context, in the context of the PCDE 3 that were relied upon for Request 14 didn't 4 contain the truncated version of the word 5 "guidelines". 6 Q. Has that been rectified? 7 A. Yes, it has. 8 Q. Now, you address in paragraphs 79 and 80 on 9 page 30 of your witness statement what in fact 10 happened when Appendix 3 was discovered and how 11 Appendix 6, which was part of the same family of 12 documents, was not discovered. If we can turn 13 that up please. So 79 first: 14 "I have sought to understand the 15 circumstances which led to Appendix 3 but not 16 Appendix 6 being disclosed to the Inquiry ... 17 and my firm's involvement. I have done so 18 having particular regard to the fact that my 19 firm reviewed Appendix 3 in context of our work 20 on this Request and therefore, on the face of 21 things, Appendix 6 was within our reach. 22 "I understand the position to be as follows. 23 A number of versions of Appendix 3 were reviewed 24 by my firm's ALT team and marked irrelevant as 25 part of the initial review of search-term 209 1 responsive documents in March 2022. It is 2 apparent from data on Relativity that none of 3 these were in families containing Appendix 6." 4 Just stopping there, what you're saying is 5 that the Appendix 3 document existed in multiple 6 versions that were reviewed by the lawyers in 7 the ALT team abroad? 8 A. Yes, not lawyers, generally law graduates, but 9 yes. 10 Q. How many versions of Appendix 3 were discovered? 11 A. On that occasion, I think nine is the right 12 number. Mr Tombleson's statement, I think, 13 explains how many there were and I think it was 14 nine. 15 Q. Were all of them reviewed by your firm's ALT 16 team? 17 A. Yes. 18 Q. Were they reviewed by different people? 19 A. Yes. 20 Q. They were all marked as "irrelevant"? 21 A. Yes. 22 Q. Then if we turn to paragraph 81, which is over 23 the page, thank you: 24 "Subsequently, as part of the further review 25 of materials in April 2022 that I describe at 210 1 paragraph 73 above ..." 2 Just cutting to the quick on that, a data 3 error had been discovered and therefore the 4 exercise was carried out again? 5 A. Yes. 6 Q. This time in London by lawyers? 7 A. Yes. 8 Q. You say: 9 "... a senior associate [at my firm] 10 reviewed a number of versions of Appendix 3, 11 some of which were in families containing 12 Appendix 6." 13 So this happened essentially by chance 14 because of the data error. 15 A. Yes, yes it did. The first search that had been 16 reviewed in March had been the subject of 17 item-level de-duplication, so my understanding, 18 de-duplicated out of the iterations of 19 Appendix 3 that were reviewed in that context 20 were versions that contained Appendix 6 as 21 a family member. But yes, you're right, the 22 subsequent review was -- it was necessitated by 23 an error. 24 Q. So if the data error hadn't happened, we 25 wouldn't even have got Appendix 3? 211 1 A. Not in response to Request 11, no. 2 Q. So the senior associate reviewed a number of 3 versions of Appendix 3. Do you know how many 4 the senior reviewer reviewed? 5 A. I don't recall. I think perhaps six or seven. 6 Again, I think it might be in Mr Tombleson's 7 statement. 8 Q. The reviewer tagged one version of Appendix 3 as 9 relevant and the document was produced to us in 10 May 2022. So here, some of these versions of 11 Appendix 3 were in families containing 12 Appendix 6, yes? 13 A. Yes. 14 Q. So here, de-duplication wasn't an issue, 15 whatever KPMG may have been doing in relation to 16 other Rule 9s or other parts of Rule 9s, so far 17 as de-duplication was concerned. That's not the 18 problem here? 19 A. No. 20 Q. Then if we look at paragraph 82: 21 "I have spoken with the reviewer ... He 22 explained he was familiar with both the 23 Prosecutions Policy and the Investigations 24 Policy ... His assessment, so far as he can 25 recall now, was he considers that Appendix 3 was 212 1 not responsive to [Rule 9(11)] ... because it 2 was not an iteration of either of those 3 policies. However, he considered that the 4 document might be of interest to the Inquiry, 5 and he tagged the document as relevant on that 6 basis. [You] believe that [this] was correct. 7 He did not look at the family documents of 8 Appendix 3, and did not review Appendix 6. In 9 fact, as far as [you are] aware, no member of 10 [your] team reviewed Appendix 6 until May 2023." 11 So you say you didn't look at the family 12 documents of which Appendix 3 was a part, why 13 not? 14 A. I think principally it was due to the approach 15 that was being taken in the context of that 16 review, which was again seeking the two specific 17 documents, not to look at families. 18 Q. But here, he wasn't, on what he's told you, 19 discovering a document that was responsive to 20 the request. He was finding a document, 21 rightly, in which the Inquiry would be 22 interested, and saying it's relevant. Why not 23 look at the family of which it's a part? 24 A. I fear I'll be repeating the answer I gave, but 25 the approach that was being taken in the context 213 1 of that review was to focus on the keyword 2 responsive documents, again the objective being 3 to find iterations of these two policies. The 4 approach that we've always taken has been where 5 we see a document incidentally that we think is 6 not responsive to a particular Rule 9 Request, 7 but we think that it will be of interest to the 8 Inquiry, that it relates to the terms of 9 reference, then we will also produce it. 10 That's what he wasn't doing -- sorry, that's 11 what he was doing here but he wasn't following 12 the train of enquiry. As I recall it, these 13 searches were conducted very shortly before we 14 gave disclosure and -- sorry, the review was 15 conducted very shortly before we gave this 16 tranche of disclosure. 17 Q. Can I turn to de-duplication. I think it's fair 18 to say that there is a dispute between your 19 firm, on the one hand, and KPMG, on the other, 20 as to who made decisions as to whether 21 de-duplication was to occur in relation to 22 a particular Rule 9 and, if so, the nature of 23 the de-duplication that was to be undertaken; is 24 that right? 25 A. I wouldn't have characterised it as a dispute. 214 1 I think it's a difference of view. In terms 2 of -- 3 Q. So there's a difference of view as to who made 4 the decision? 5 A. In relation to the latter part of your question, 6 yes. No question on -- there's no disagreement 7 as to the occasions on which de-duplication was 8 instructed. The difference of view relates to 9 the nature of the de-duplication that ought to 10 have been applied. 11 Q. Would you agree that it's unsatisfactory, to say 12 the least, that there is a difference of view, 13 using your language, between two of the Post 14 Office's service suppliers on such a critical 15 issue as this, which has had such consequences 16 for both the Post Office and the efficient and 17 effective conduct of this Inquiry? 18 A. Mr Beer, we have different views on a point. 19 I think -- 20 Q. Do you think it's a good thing? 21 A. No, it's not a good thing. It's regrettable. 22 Q. I'm sorry? 23 A. I say it's regrettable. 24 Q. My phase was "unsatisfactory". Are you willing 25 to go as far as to say that you think it's 215 1 unsatisfactory? 2 A. No, I'm not, actually, I think, because the 3 views are genuinely held, and it's a regrettable 4 situation. 5 Q. Who made the decision to carry out item-level 6 de-duplication? 7 A. I believe that KPMG made the decision. 8 Q. And why do you believe that? 9 A. Because I have reviewed some of the instructions 10 that members of my team gave to KPMG and the 11 back and forth that occurred with KPMG in 12 relation to the application of de-duplication, 13 and I don't think that the instructions that 14 were given can fairly be characterised as 15 instructions to carry out item-level 16 de-duplication. 17 Q. We have just heard evidence from Mr Tombleson, 18 who said he'd done pretty much the same exercise 19 within KPMG, that KPMG undertook searches based 20 on search terms provided by HSF, provided 21 statistics as to the number of responsive hits. 22 There was then liaison between HSF and KPMG in 23 which, in some respects, HSF said the number of 24 hits seemed high, or too high. And then a table 25 was given, if de-duplication was applied at an 216 1 item level, although that phrase was not used. 2 And then instructions came back from your 3 associates or senior associates saying yes, 4 carry out the de-duplication. Is that a fair 5 summary of the correspondence that you've read? 6 A. I would describe it differently. I think, for 7 one thing, there are many different email chains 8 with instructions made, given in different 9 contexts, from different people to different 10 people, with different terminology used. 11 I think what's clear is that members of my team 12 did not wish to review multiple iterations of 13 the same keyword responsive document. If 14 they're reviewing one iteration of a document, 15 it's not necessary to review 15 others. So they 16 didn't want -- 17 Q. That sounds reasonable. 18 A. Well, quite, and I think no one would argue with 19 that. That was the objective that was being 20 articulated in various different ways in the 21 instructions that were given to KPMG. My 22 understanding -- and I should say that I have no 23 expertise in these issues, but I'm familiar with 24 eDiscovery from having practised for many years, 25 and I've also discussed these issues with my own 217 1 firm's eDiscovery specialists. In those 2 circumstances, there are a number of different 3 options that might be pursued, of which 4 item-level de-duplication is the most extreme, 5 and very extreme. And that was the option that 6 was pursued. 7 Q. I mean, if you Google "item-level 8 de-duplication" and "MD5#", you can see quite 9 quickly that it's rather deprecated. 10 A. Yes. 11 Q. I think the first three hits you get tell you 12 it's rarely used and can have disastrous 13 consequences. Did you know that? 14 A. I did when I ran the same Google search as you, 15 evidently have done. 16 Q. But is the problem here that you're saying that 17 you and your firm didn't know that it was being 18 done in your firm's name -- 19 A. Yes. 20 Q. -- or the Post Office's name? 21 A. Yes, I think it is. I think the level of 22 understanding of the different forms of 23 de-duplication and the different techniques that 24 one can apply to achieve the objective 25 I described, that will differ amongst different 218 1 lawyers in the team, depending on their level of 2 experience, whether they've happened upon this 3 before. I, for my part, I don't think it can be 4 fairly expected of solicitors that they would be 5 familiar with different forms of -- different 6 eDiscovery techniques for achieving that 7 objective. 8 Q. Did KPMG say in any of the communications that 9 you've reviewed "There are two types or at least 10 two types of de-duplication. If you do 11 item-level de-duplication, you'll get a much 12 smaller number of documents that are responsive, 13 fewer hits, which is good, but you will occlude 14 from view a very large number of family 15 documents"? 16 A. If I may, I don't think it's so much just 17 removing the documents from view; it's the 18 actual -- it's the exclusion of those documents 19 from the dataset in a way that makes them, as 20 we've seen, very difficult to retrieve and to go 21 back to the position before item-level 22 de-duplication was applied. So it's that 23 consequence, that's the particularly severe 24 consequence, or one of them, of item-level 25 de-duplication -- 219 1 Q. Well, it's a severe consequence if you're trying 2 to do a remediation exercise. For the Inquiry, 3 the severe consequence is that those documents 4 aren't reviewed for disclosure, the 5 de-duplicated families -- 6 A. Forgive me. 7 Q. If you apply item-level de-duplication -- 8 A. Yes. 9 Q. -- the other documents that are part of the 10 families are occluded from view. They're not 11 part of the search -- 12 A. Yes. 13 Q. -- and therefore can't be part of a review, and 14 therefore can't be disclosed to this Inquiry in 15 response to a Rule 9 Request. That's the 16 principal vice, isn't it, of an item-level 17 de-duplication exercise? 18 A. It's an aspect, but the point I was trying to 19 make was that another significant aspect is the 20 difficulty with getting the documents back. And 21 putting that crudely, there's more complexity -- 22 Q. That's only an issue if you realise the mistake 23 later down the line and you have to scrabble 24 around and do a remediation exercise, isn't it? 25 A. No, I wouldn't agree with that. It might be 220 1 an issue a few days later. You may review the 2 documents set that has been made available 3 following the de-duplication having been run, 4 and you may take the view that you want to 5 review more documents. 6 Q. But in any event, KPMG didn't say, in any of the 7 communications, "There are these different 8 species of de-duplication, and they have these 9 consequences, which would you prefer, HSF?" 10 A. No, they didn't. There was an email in June 11 this year explaining the -- setting out the 12 different types of de-duplication. 13 Q. Which is all a bit late, isn't it? 14 A. Yes. 15 Q. You say about this in paragraph 121(j), which is 16 on page 48 of your witness statement, (j). 17 That's it: 18 "As to the impact of de-duplication more 19 generally, I've noted ... above that there 20 appears to [be] a misalignment between my team 21 and KPMG as to the approach to de-duplication 22 and its implications." 23 That's a rather enigmatic phrase, if I may 24 say so, "a misalignment as to approach". Are 25 you describing, by the use of the phrase 221 1 "misalignment", a misunderstanding between those 2 giving the instructions and those receiving 3 them? 4 A. Yes, in part, I am. 5 Q. Can I turn to remediation work, then, please. 6 If we can turn up in your witness statement 7 page 64, please. This is the start of the annex 8 to your witness statement; is that right, 9 Mr Rowan? 10 A. Yes, it is. 11 Q. If you go to the foot of the page, please, 12 you'll see that the annex is 136 pages long. 13 You've kindly set out for us in detail the 14 remediation exercise that has been undertaken in 15 relation to the three issues that we have been 16 discussing but, additionally, a disclosure 17 assurance exercise that has been undertaken in 18 conjunction with the Post Office; is that right? 19 A. Yes. 20 Q. I'm not going to go through any of that, it's 21 there for anyone that's interested to read it 22 but, just in relation to the timescales of some 23 of it and the implications for Phase 4 of the 24 Inquiry, which starts in a fortnight -- or it 25 continues in a fortnight -- it's right, isn't 222 1 it, that the chairman directed that disclosure 2 for witnesses commencing their evidence on 3 19 September this year and following must be 4 complete by 14 August 2023. They were his 5 directions? 6 A. Yes. 7 Q. So that gave the Inquiry a little over a month 8 to digest the disclosure, to analyse it, to 9 disclose it to Core Participants, to allow Core 10 Participants to digest and to analyse it and to 11 propose questions to upcoming witnesses? 12 A. In relation to these -- 13 Q. The first witnesses? 14 A. Yes, the first witnesses in the case study, yes. 15 Q. It appears from the annex and updates to the 16 annex which we have received recently that that 17 deadline was not met and is not going to be met 18 by the Post Office; is that right? 19 A. No, the work is ongoing. 20 Q. Instead, a principal deadline of 14 September 21 has been set -- 22 A. Yes. 23 Q. -- by the Post Office itself, or a deadline that 24 it is working to? 25 A. I wouldn't call it a deadline, as such. I'd 223 1 call it a realistically achievable target date 2 that the entire teams -- my own, Peters & 3 Peters, Post Office -- are working very hard to 4 meet. 5 Q. So that's Thursday, 14 September, with us 6 commencing our witnesses on Tuesday, 7 19 September, giving us the two working days to 8 look at any material that arrives on the 14th? 9 A. The way we have approached the exercise is to 10 try to prioritise by reference to the Inquiry's 11 running order for the rest of the year. So in 12 relation to -- I think I'm right in saying it's 13 the case study that comes first -- that was 14 prioritised and closed out some time ago. So my 15 hope is -- and there was a relatively small 16 number of documents produced in relation to 17 that, so my hope is that, in relation to the 18 case study, that would allow sufficient time for 19 the documents to be reviewed, and certainly more 20 than a couple of days. 21 Q. In relation to the documents that have been 22 disclosed as a result of the remediation 23 exercise, I think you've kindly prepared a table 24 for us. 25 A. Yes. 224 1 Q. I'm going to work over the one that was 2 disclosed last night rather than the one at 8.21 3 this morning. Is it right that the total number 4 of documents produced to the Inquiry as a result 5 of the remediation exercise is 7,753? 6 A. Yes, it is. 7 Q. That includes a number of duplicates and near 8 duplicates -- 9 A. Yes. 10 Q. -- despite the remediation of the de-duplication 11 errors? 12 A. Yes. 13 Q. Could you explain, please, why a re-duplication 14 exercise has been undertaken only for duplicates 15 and near duplicates then to be disclosed to the 16 Inquiry? 17 A. The exercise of remediating item-level 18 de-duplication has involved reviewing the 19 documents that were excluded by item-level 20 de-duplication, which would have included, as 21 we've discussed, duplicates of keyword 22 responsive documents and their families. So 23 a significant part of the production, certainly 24 the production that's to remediate item-level 25 de-duplication, will contain, to that extent, 225 1 duplicate documents, albeit they sit in 2 different families. 3 More generally, by duplicate, the way we've 4 actually characterised duplicates here is to 5 talk about duplicates or near duplicates. 6 Relativity has functionality that enables 7 documents that don't have the same MD5# value, 8 but which are nevertheless substantively the 9 same or substantively very similar, to be 10 identified. So what we have done is used that 11 functionality, to run that over the documents 12 that have been produced and to draw to the 13 Inquiry's attention the extent of the near 14 duplicacy that exists in the document set, which 15 is very significant. 16 MR BEER: Thank you very much. 17 Sir, they're all the questions that I ask of 18 Mr Rowan. 19 Oh, I'm just having something shown to me. 20 They're all the questions I ask of Mr Rowan. 21 Do you have any questions? 22 SIR WYN WILLIAMS: No, thank you. 23 So that concludes the evidence that we set 24 out to hear today. 25 Thank you, Mr Rowan, for your very detailed 226 1 evidence and for the detailed answers you've 2 given to Mr Beer. 3 I don't propose to say anything at all this 4 afternoon about the evidence which I've heard 5 during the course of today or, indeed, to make 6 any kind of comment about what I intend to do by 7 way of making known my views, if any, upon that 8 evidence. I want to reflect on what I do over 9 the next day or two and then I will, at that 10 stage, make a short statement in writing as to 11 how I intend to proceed. 12 MR BEER: Thank you very much, sir. That aside, we 13 next see each other, I think, on Tuesday, 14 19 September -- 15 SIR WYN WILLIAMS: That is true. 16 MR BEER: -- albeit there may be a hearing in the 17 week before. 18 SIR WYN WILLIAMS: Yes. There is the possibility -- 19 let me use that word as neutrally as possible -- 20 of a hearing, I believe, on Thursday, 21 14 September. Is that the correct date? 22 MR BEER: It is, sir. But you've not taken 23 a decision on that because you're waiting for 24 some submissions on that discrete issue. 25 SIR WYN WILLIAMS: Exactly so. So again, I say no 227 1 more about that possibility. Right, that's 2 fine. 3 So thank you all for participating and, 4 unless we meet on 14 September, I will see you 5 all on 19 September. 6 MR BEER: Thank you very much, sir. 7 (4.36 pm) 8 (The hearing adjourned until a future date) 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 228 I N D E X FINTAN CANAVAN (affirmed) .....................3 Questioned by MR BLAKE ........................3 Questioned by SIR WYN WILLIAMS ...............46 DIANE WILLS (sworn) ..........................47 Questioned by MR BEER ........................47 PAUL JOHN TOMBLESON (sworn) .................131 Questioned by MR BLAKE ......................131 GREGG NICHOLAS ROWAN (sworn) ................164 Questioned by MR BEER .......................164 229