The efforts made to resist providing metadata
The preceding account of Dr Wright’s activity on his Overleaf account has only emerged as a result of documents and information provided by Shoosmiths during the trial.
Dr Wright had numerous opportunities to tell the truth – that the so-called Bitcoin White Paper LaTeX Files were the product of days of work done on Overleaf in November 2023 and not the processes described in his first and fourth witness statements but he failed to take them:
He said nothing to that effect when asking for an adjournment of the trial at the PTR.
Nor did he provide any such description when, on 20 December 2023, he served Wright8, a 24-page statement in which he purported to provide details of his LaTeX environment. {E/23}.
Nor did he provide that description on 12 January 2024, when he served his reply evidence, Wright11. Shoosmiths’ letter of 13 December 2023 stated that this statement would give Dr Wright’s account of “the corrections he has made to the White Paper LaTeX files since the first publication of the Bitcoin White Paper, to the best of his recollection given the passage of time”. His witness statement gave no such account. In truth, there had been no “corrections” to the files, and the passage of time (less than two months) was unlikely to have clouded Dr Wright’s recollection.
Nor did he provide any such description in Wright14, served on 30 January 2024 and produced in response my Order that Dr Wright identify the chain of custody in relation to the so-called White Paper LaTeX Files {E/33}.
As an aside, Dr Wright did make reference to how the files had been stored on the QNAP server in Wright14. At one point it seemed that Dr Wright might place some emphasis on this server as a repository of relevant information. In Wright14 at {E/33/4} Dr Wright explained how Alix Partners came to copy the QNAP server and took it away. Dr Wright suggests that he copied the White Paper LaTeX Files onto an external drive from the QNAP server at that time. I agree that suggestion cannot be true. AlixPartners have confirmed that they collected the QNAP server on 4-5 February 2019: see Shoosmiths’ letter of 29 January 2024 at {M1/2/138}. AlixPartners inspected the QNAP server onsite, detected it was encrypted and took it away {M1/2/139}. They did not seek to image the QNAP server because it was encrypted, as Dr Wright has confirmed {M1/2/140}. Given that the QNAP server was “not accessible to them without valid credentials and keys” and “inaccessible whether by AlixPartners, [Dr Wright] or a third party” {M1/2/140}, Dr Wright cannot have accessed it to remove the White Paper LaTeX Files. In Tulip Trading, Dr Wright has suggested that the QNAP server is not even owned by him, but is instead owned by nChain {S1/1.36/2} at [5] and so the QNAP server faded from attention in the present proceedings.
Dr Wright dragged his feet in providing any useful metadata in compliance with my Order made at the PTR:
By January 2024 no metadata had been provided – and Dr Wright had not even provided Shoosmiths with the login credentials to his Overleaf account {Shoosmiths’ letter of 8 January 2024 {AB/2/189} at [1]}.
Accordingly, COPA wrote to Overleaf directly on 3 January 2024. Overleaf responded to say that they had provided information to Shoosmiths, and separately wrote to Shoosmiths directing them to the Project History feature on Dr Wright’s Overleaf account.
On 8 January 2024, Shoosmiths provided 17 files said to have been provided by Dr Wright to demonstrate Overleaf’s “Other logs and files” feature. None of those files was at all informative as to Dr Wright’s activity on the account.
The Developers wrote to Overleaf on 10 January 2024 requesting that they produce the relevant files. That same day, Shoosmiths wrote to advise that Overleaf had emailed to Dr Wright “an export of the project history” for his account. In the event, Overleaf declined to provide information to the Developers.
Faced with the imminent start of trial, the Developers made an application for Dr Wright to be ordered to consent to Overleaf providing the data from his account on 16 January 2024. That application prompted Shoosmiths to confirm that they would be “in a position to provide the materials requested” on 22 January 2024.
On 22 January 2024, Shoosmiths finally produced four zip files, including one containing a redacted version of the data that had been supplied by Overleaf. Even now it is not clear who had decided on the relevant redactions or why. The files included:
A file entitled REDACTED_project.json, which related to the Maths (OLD) folder, the existence of which had not previously been revealed. The file showed that Dr Wright had carried material from the Maths (OLD) project into the (disclosed) main.tex file of the Bitcoin project.
A spreadsheet entitled “chunks” that had been prepared for the Bitcoin project. This file recorded the changes made to the main.tex file in the ‘Bitcoin’ folder set out at paragraph 455 above.
On 1 February 2024, after being pressed further in correspondence, Shoosmiths wrote to Macfarlanes about the Maths (OLD) project. They confirmed that Dr Wright had put the REDACTED_project.json file associated with the Maths (OLD) project into the Bitcoin folder that had been disclosed to COPA and the Developers “inadvertently”. As Counsel submitted, put another way, Dr Wright had never intended to reveal the existence of the Maths (OLD) project to the Developers.
However, the materials disclosed from the Maths (OLD) project were a revelation. They showed for the first time details of Dr Wright’s repeated tinkering with the White Paper LaTeX files between 17 and 19 November. Nevertheless, the Maths (OLD) files were defective in two respects:
First, Dr Wright was continuing to assert privilege over some of the files: including the ZZZ and Test subfolder into which Dr Wright had first placed the BitcoinSN.tex file {see {M1/2/153} at [2.d]};
Second, the changes to the White Paper LaTeX files were shown in a Maths (OLD)_chunks spreadsheet rather than in their native chunks.json format. As a result (although it was possible to track most of the changes on a row-by-row basis from the spreadsheet) it was impossible to rebuild and compile the changes sequentially from the available data.
On Day 5 of the trial, Mr Hough KC turned to the topic of the White Paper LaTeX Files in his first cross-examination of Dr Wright. In particular, he took Dr Wright to the chart set out at paragraph 455 above and suggested that Dr Wright was responsible for the edits shown in that document. Dr Wright admitted that he was, but then said that he had made all of the changes during demonstrations to Shoosmiths:
“153: 5 Q. You were responsible for those edits, weren't you,
6 Dr Wright?
7 A. I was.
8 Q. So the file was being edited right up to the day before
9 the LaTeX files were received by Stroz Friedberg?
10 A. Yes. I demonstrated to Shoosmiths, making a small
11 change, adding a full stop, adding a percentage. And
12 where you say there are extensive edits, that's actually
13 not true. Adding a full stop, removing that full stop,
14 is actually two edits. So, when I add a space, that's
15 an edit. If I go percent, comma, slash, etc, that's
16 three edits. So, at one stage, I typed in Matt's, one
17 of my solicitor's, name. That was probably 10 edits.
18 I then undid it and put the original name back. So
19 I was demonstrating how using that, you could change
20 the date and produce a new version, etc.
21 Q. Dr Wright, first of all, this was a document which you
22 were going to present as being a perfect digital
23 watermark of the Bitcoin White Paper. Didn't it occur
24 to you, as an IT security expert, that you shouldn't be
25 mucking with it extensively over the period of time
154: 1 before you produced it?
2 A. I downloaded a copy of the file and gave it to
3 Shoosmiths before I did any of this. So, the first
4 thing is, I downloaded the ZIP from Overleaf, sent it to
5 the solicitors. We did that right at the beginning of
6 this process. And as such, once I've given them a copy,
7 I'm saying that I can't change the copy they have,
8 therefore my making changes and undoing those changes is
9 not a material change.
10 Q. Do you say that all of those edits were done in
11 the presence of Shoosmiths?
12 A. They were on videos, on calls, I sent them some emails
13 while they weren't on there, I sent, like --
14 Q. You say that all these edits were done in their
15 presence?
16 A. Not in their presence. I emailed them. They weren't
17 there. And do you consider on a video call presence?”
Although Dr Wright’s Counsel then observed that issues of privilege were being traversed, Dr Wright’s answers put Shoosmiths in an impossible position. I agree that it plainly was not true that Dr Wright had made the changes in demonstrations to Shoosmiths. Indeed, as noted at paragraphs 453 to 454 above, he had not even mentioned any changes in his emails to Shoosmiths enclosing the Bitcoin folder.
In these circumstances, there had to be a waiver of privilege in relation to the Maths (OLD) and Bitcoin folders – and that occurred on Friday, 16 February 2024. For the first time, unredacted chunks.json files were produced by Dr Wright. That enabled the Developers to set about compiling each of the revisions that Dr Wright had made to the White Paper LaTeX Files between 17 November 2023 and 24 November 2023. The Developers presented that work to Shoosmiths on Monday 19 February 2024, together with the code that they had developed to compile the documents from the chunks.json files.