Anthropic has just announced that Claude will start leaving an invisible mark in the text it generates, in an announcement that may have unintended consequences for expert witnesses.

Supported Claude models will embed machine-readable watermarks into generated text. Anthropic says those marks will travel with the words when copied and pasted into other documents and may survive subsequent editing.

The immediate purpose is provenance. As the volume of AI-generated material increases, watermarking offers a potential means of identifying content that has passed through a particular AI system.

For expert witnesses, however, that raises a rather more difficult question: if an expert report contains an AI watermark, what exactly does that prove?

Perhaps less than might first appear.

Detection is not the same as authorship

One of the most important qualifications comes from Anthropic itself.

Anthropic cautions that detecting its watermark does not establish that Claude wrote or originated the material. Rather, it indicates that the content may have been processed by Claude.

That distinction could be particularly important in expert evidence.

An expert might write an entire report themselves and subsequently ask an AI tool to correct spelling and grammar. Another might use AI to improve the structure of material they have already written. A third might provide their analysis and ask an AI system to turn it into polished prose.

Those activities are quite different from asking an AI system to examine the underlying evidence, perform the analysis and suggest the opinion the expert should express.

Yet all involve AI-generated or AI-processed text.

Consider four possible uses:

• An expert writes their report and uses AI to check spelling and grammar.
• The expert writes the substance but asks AI to improve its structure or expression.
• The expert provides their analysis and asks AI to draft passages expressing it.
• The expert asks AI to analyse the evidence and assist in formulating the opinion.

All involve AI, but they plainly do not raise the same questions about the independence of the expert’s analysis and opinion.

That makes the presence of a watermark potentially useful evidence, but not necessarily evidence of what might initially be assumed.

Provenance of the text is not provenance of the opinion

This is perhaps the central issue for expert witnesses. A watermark may provide evidence about the provenance of the words. It might indicate that Claude, or another AI system using similar technology, has been involved in generating or processing them:

• It cannot, by itself, establish the provenance of the opinion.
• It does not tell us who examined the evidence.
• It does not tell us who identified the relevant issues.
• It does not tell us whether the expert carried out the underlying calculations or analysis.
• And, most importantly, it does not tell us whether the conclusion was independently reached by the expert or suggested by an AI system.

The same paragraph could conceivably carry an AI watermark whether Claude corrected three grammatical errors in it or generated the paragraph from scratch.

For expert evidence, those are very different propositions.

The absence of a watermark proves little too

There is also a problem in the opposite direction.

No watermark does not necessarily mean no AI.

Anthropic acknowledges that detection can become more difficult where generated material has subsequently been heavily edited, paraphrased, translated or combined with other text. A passage may also simply be too short to provide sufficient information for reliable detection.

This creates an important limitation if watermark detection ever begins to be used when examining expert reports.

A positive result may indicate AI involvement without establishing the extent or nature of that involvement whilst a negative result cannot necessarily establish that there was no AI involvement at all.

Watermark detection should therefore not be treated as a binary test of whether an expert “used AI”.

A new question in cross-examination?

The development also raises an interesting practical possibility.

“Your report contains a watermark indicating that it was processed by Claude. How did you use it?”

The answer might be entirely unremarkable.

The expert might explain that they wrote the report themselves and used Claude to check its grammar. They might explain that they used it to condense a lengthy passage they had already written. Alternatively, the answer might reveal that AI played a much greater role in analysing material or formulating the opinions contained in the report.

The takeaway is that whilst the watermark itself cannot distinguish between those situations, it may, however, provide the starting point for questions which can.

That means the increasingly important question may not simply be:

“Did you use AI?”

It may instead be:

“Precisely how did you use it?”

Watermarking is becoming a real technology

Anthropic is not alone in pursuing this approach.

Google has already developed SynthID Text, a watermarking technology for AI-generated text. Rather than inserting a visible label or conventional metadata, SynthID subtly affects the model’s choice of tokens as text is generated. Across a sufficiently long passage, those choices create a statistical signature which can subsequently be detected.

Anthropic has not publicly confirmed that Claude’s watermark uses the same technical method, so the two systems should not be assumed to work identically.

The significance is broader. Text watermarking is moving beyond academic research and into mainstream generative AI products.

If the technology becomes widespread, lawyers, courts and experts may increasingly encounter claims that a document has been identified as having passed through an AI system.

Understanding what such a finding does, and does not, establish will therefore become important.

The expert remains responsible

The Academy’s guidance on the use of Artificial Intelligence by expert witnesses already addresses the more fundamental issue.

AI can be a useful tool, but it cannot substitute for the expert’s own expertise, analysis and opinion. Experts remain responsible for the evidence they give. Where AI has played a significant role, keeping appropriate records of how it was used may also become increasingly important, particularly if the expert is subsequently required to explain that use in court.

Watermarking adds another dimension to that responsibility.

It may make some uses of AI more readily identifiable. But identifying that an AI system touched the words is not the same as establishing who produced the intellectual work behind them.

For expert evidence, that distinction is crucial.

Provenance of the text is not necessarily provenance of the opinion.

As AI provenance technology develops, the ability of an expert to explain what the AI did, what the expert did, and where the opinion actually came from may become increasingly important.

In Trinsic Collagen Ltd v Procoll Ltd [2026] EWHC 1793 (Ch), the court criticised the way additional information had been provided to the claimant’s experts and then relied on without being clearly identified in the reports, or in one instance the joint statement.

The issue was not that the experts had spoken to the client. It was that the factual basis of their conclusions was not always transparent.

For an expert report to assist the court properly, the reader should be able to see the route from the facts and assumptions, through the analysis, to the opinion reached. That means being clear about material obtained through discussions, site visits, correspondence or subsequent instructions, especially where it influences the expert’s reasoning.

The Academy’s own guidance reflects the same principle. An expert report should contain a concise statement of the facts and assumptions used, followed by the expert’s analysis and opinion, with fact and opinion clearly distinguished.

A useful question before signing a report is this: could another expert, or the judge, identify every material fact and assumption on which my conclusion depends? If not, the report may need another look.

The Academy’s Model Form of Expert Report also reflects this approach.

Not later. In the report.

If an expert opinion depends on incomplete data, assumptions or qualifications, those limitations belong in the written report.

1. Not in cross-examination.
2. Not in re-examination.
3. Not only after the tribunal asks the right question.

In Mew v General Dental Council [2026] EWHC 1116 (Admin), the High Court considered expert evidence given during professional disciplinary proceedings.

One expert had deliberately avoided setting out limitations in the available data, believing that they could instead be explained when he gave oral evidence.

Charles Bagot KC, sitting as a Deputy High Court Judge, said that this revealed a “fundamental misunderstanding” of the obligations of a CPR-compliant expert.

The problem was not that the evidence required qualification. Many expert opinions do.

The problem was that the qualification had been withheld from the report.

A court or tribunal must be able to understand:

• what the expert can conclude;
• what the expert cannot conclude;
• where the evidence is uncertain; and
• how those limitations affect the strength of the opinion.

A carefully expressed qualification does not weaken expert evidence. It allows the decision-maker to assess it properly – and will often make the opinion more credible.

The expert report should contain the opinion the expert can genuinely support, not an unqualified version that appears more persuasive on first reading.

Read the judgment:
https://www.bailii.org/ew/cases/EWHC/Admin/2026/1116.html

“The expert had only skimmed the papers.”

Few expert witnesses would want to read those words in a judgment.

In Re X and Y (Care Proceedings: Publication of Judgment: Criticism of Expert Witness) [2026] EWFC 132, Mrs Justice Knowles decided that the consultant neurosurgeon should be identified following serious criticism of his conduct as an expert witness. That decision followed the court’s detailed findings in Re X and Y (Care Proceedings: Fact Finding: Death of a Child: Expert Evidence) [2025] EWFC 132.

The criticism went far beyond inadequate preparation.

The court found that the expert had not properly engaged with the available evidence, had strayed beyond the limits of his expertise, had failed to engage adequately with the opinions of other experts, and had advanced conclusions based on speculation rather than a sound evidential foundation.

The case is a reminder that expert evidence is judged not only by the conclusions reached, but by the process used to reach them.

An expert opinion carries weight because it is the product of careful preparation, objective analysis and transparent reasoning.

Before accepting an instruction—or before entering the witness box—it is worth asking:

✔️ Have I fully engaged with all the relevant material?

✔️ Am I remaining firmly within my own expertise?

✔️ Have I fairly considered competing expert opinions?

✔️ Can I clearly explain the evidential basis for every conclusion I have reached?

The role of an expert witness is not to persuade. It is to assist the court with independent, balanced and properly reasoned evidence.

📖 Judgments:

Re X and Y (Care Proceedings: Publication of Judgment: Criticism of Expert Witness) [2026] EWFC 132
https://www.bailii.org/ew/cases/EWFC/HCJ/2026/132.html

Re X and Y (Care Proceedings: Fact Finding: Death of a Child: Expert Evidence) [2025] EWFC 132
https://www.bailii.org/ew/cases/EWFC/HCJ/2025/132.html

 

Some specialist areas of expertise are so niche that it’s almost impossible to find experts who are free of previous involvement with one or more parties to a dispute. In a recent case requiring expert evidence in mine surveying (Cleveland Potash Ltd v Drummond & Ors) not only had both experts been involved with the parties they had worked in the same office as each other for 20 years!

In this case both experts were clear and transparent about their previous involvement and as the  judge explained, “Whilst neither party raised any concerns about the experts’ prior and ongoing involvement in the issues related to this dispute I will have to consider that involvement as part of my overall assessment of the experts. I remind myself that the experts have an overriding duty to the court which should override their duty to their clients.”

It seems that one expert was much more successful in retaining her objectivity than the other, reading the judge’s explanation gives some important pointers for all experts as to how to do this, specifically;

“Mrs Iwanicki was a very impressive witness. She was deeply knowledgeable about her subject. She gave clear, careful, well considered and measured answers to the questions she was asked. She had a good understanding of her overriding duty to assist the court rather than to advocate for a particular position. Her answers were honest and straightforward even where those answers might have been seen as providing some assistance to Rs. That is what the court expects of an expert witness who understands that their role is not to advocate for a particular party or outcome but to assist the court.

 

Mr Troman was also clearly very knowledgeable about his subject, however, unlike Mrs Iwanicki on some occasions he did not appear able to disassociate his longstanding role of assisting the Rs (and other leaseholders) to negotiate what he considered to be a satisfactory outcome with CPL from his duty to assist the court. This more partisan approach had infected some of his evidence and the contents of his report. It was clear from his evidence that he felt that other mineral rights owners who had now agreed terms with CPL should have held out and/or would have done better if they had followed his advice/recommendations. He was reluctant to accept the possibility that the conclusions in his report might need to be adjusted or modified even when it became apparent that there were factors or information that he had not known about and could not therefore have taken into account when reaching his conclusions. His resistance to accepting any possibility that his conclusions on valuation might need to be adjusted was unhelpful and undermined his position. Further he had a tendency to speculate both in his report and in his evidence and on at least one occasion asserted a factual position which turned out not to be true. Overall, this approach undermined his credibility and the weight that I give his expert opinion.

 

As always the key for any expert is to ask yourself ‘what would I say if I was acting for the other side?’. If the answer is in any way different then you need to think again. For guidance on what to do if you have been involved with a party to a dispute and other issues relating to conflicts or perceived conflicts see our guidance on the subject.

Our image is from the New Scientist

 

The Pilot for Access to Public Domain Documents (governed by Practice Direction 51ZH) was originally intended to commence in October 2025 but will now go live from 1^st January 2026.

The pilot will apply to the Commercial Court (including the London Circuit Commercial Court) and the Financial List (except where a party is unrepresented and had not used the Court’s CE File electronic filing system previously). It is intended that there will be a review after 6 months although the pilot is not scheduled to finish until 31^st December 2027.

If the pilot is successful, it will then be extended to other courts – most likely the Business and Property courts first.

The Pilot identifies certain documents as Public Domain Documents. It is intended to only capture documents which as a matter of law enter the public domain when a hearing takes place in public without imposing an undue burden on the parties or the Court. The Pilot will see a change from the current position, whereby an application is required to access the documents, to the default being the provision of the documents.

The main categories of documents covered by the Pilot are:

• Written submissions – skeleton argument and written and opening closing submissions, as well as all other submissions provided to the judge during the course of the hearing
• Witness statements/affidavits
• Expert reports (including annexes and exhibits)

How does it work?
The documents will be accessed via CE-File which is the court’s existing system with the parties filing the documents themselves in the appropriate category.

There will be a default filing period for documents. This will be from the day they enter the public domain by being referred to in a court hearing and ending at 16:00 14 days later unless otherwise ordered or agreed by the parties.

As filing will be the default position any party who wishes to limit or not file will need to seek an order to that effect before the filing deadline (Filing Modification Orders – FMOs).

The PD contains a provision for an order to be made compelling parties to make the filing and could, if not complied with, result in appropriate sanctions such as contempt of court.

It will be the responsibility of the lawyers to upload the documents. Further information can be found in the accompanying Guidance Note.

Things to Note:
The most important thing is to be aware that if you are acting in a case which comes under the Pilot then your report will become a Public Domain Document once your evidence is given in open court. If the hearing is private, then the report will not become a Public Domain Document.

All Experts should read and familiarise themselves with both the Practice Direction and the Guidance.

Although the current scope of the Pilot is limited it is likely that in the future this will apply across the board in the civil courts and your reports will automatically become Public Domain Documents.

You should:

• Check with the instructing party whether the case will be dealt with under the pilot.
• Discuss with the instructing party when your report will be first used or referred to. It is likely that this will be when you give evidence in chief and this will be when it becomes publicly available.
• Discuss with your instructing party about how any sensitive information might be used in your report and how they would like it dealt with.
• Do not include your own personal information, other than that which is required by the procedural rules.  For example, consider not including email or telephone numbers. Remember to use your business address and not your personal one. It maybe that you consider having a separate registered business address if you work from home.
• Consider your personal media profile and the information that is publicly available – especially if it is likely that your report could led to harassment or social media comments.

The Committee intends to review the Pilot after 6 months of operation. We would appreciate any details of your experiences with the Pilot to inform our engagement with the Committee’s review.

The Academy’s GDPR guidance turned 5 this year so we thought it was a good time to review it in the light of a better understanding of how GDPR has been implemented in practice. One of the challenges of developing guidance for our expert members is they cover a wide range of specialisms and as a result deal with widely differing types of personal information. The material needed by a medical or psychiatric expert to assess a personal injury is clearly very different from that used by an accountant to assess the value of a business.

In addition while some experts work within large consulting firms many are micro-businesses who don’t have the benefit of in-house egal advice but still have to negotiate the (sometimes onerous)  privacy responsibilities passed down to them by instructing solicitors.

 

Our guidance is designed to help members consider for themselves what aspects of their work are covered by GDPR in a structured way and is linked to a model privacy policy that can be used in a ‘pick and mix’ way to include only those sections that are relevant.  We recommend micro-business or individual expert read the ICO’s small business guide alongside our guidance and remind you that you are almost certainly required to register and pay the required annual fee.

 

The Academy  provides a range of practical advice and guidance on other topics such as conficts, Joint Statements, privilege and so on which is available at no charge to members in the Academy’s Knowledge Hub.

Members can also call our Technical Helpline for help and advice on ethical and practical issues from dealing with instructions to getting paid.

If you would like to join the Academy please complete our Pre-Application Questionnaire and we’ll help you identify the appropriate membership level for you. Other benefits of membership regular member meetings to discuss issues affecting experts and access to specialised PI insurance.

The Academy held its Annual General Meeting on 14 October. As with most similar organisations, this is an important element of the Academy’s governance. Several decisions are reserved to the membership in general meeting. These include the appointment of new members of council, the election of officers and the adoption of accounts.

The following officers were elected:

• Chair: William Hooper
• Deputy Chair: Nigel Young
• Vice Chair: Mike Allen (appointed)
• Secretary: Stuart Broom
• Treasurer: Joe Skilton.

The following were elected to join the Council:

• David Cannell
• Helen Gregory
• Derek Nelson
• Syed Turab

Departing members of Council who stood down were thanked for their contributions.

The full Council meets twice a year, in June and November, and the Executive Committee once a month.

If you would like to help with the running of the Academy or stand for election to Council please contact us, volunteers are always welcome!

 

 

The Academy is the professional society and accrediting body for expert witnesses of all disciplines, independently run by experts for experts and those using them.

We’re proud to have  Lord Neuberger of Abbotsbury as our President with support from a range of legal luminaries with Chief Executive Nicola Cohen running the team who support members on a day to day basis.

But the  great strength of the Academy is the contribution made by its members to the direction and running of the organisation through the Executive who direct management & finance and Council who set  the agenda for the future and provide expert-led input into our guidance and activities.

Our AGM will be on 14^th October 2025 and this online event will bring the opportunity for members to elect their representatives to Council for next year.  If you’re a member and haven’t received an invitation please contact the Academy office.

Council meets formally twice a year and we have working groups for subjects as diverse as marketing, standards, formal consultations and AI. Current members include architects, accountants, insurance & IT experts. We’re keen to hear from members in as wide a range of specialisms as possible so whatever your interests please get in touch.

Council Members are elected for a three year term and may hold office for no more than two consecutive terms. Details of the roles and duties are available here.

If you would like to stand for election please complete the online Nomination Form and Election Address by noon on 3rd September. Please note that only Full Members and Fellows are eligible to stand for Council but Associate Members who are interested in contributing are welcome to get in touch, there is always something you can do to help!

Fans of the late, great Steve Wright will remember the amusing and amazing factoids that peppered his afternoon radio show. But ‘factoid’ also means something that is repeated or believed but isn’t actually a fact.

Experts of course need to make sure they only deal with established facts and make it clear when there are disagreements or doubt over the status of the assumed facts that underlie their evidence.

Two recent cases have pointed up the need for care in this area.  In Toppan Holdings Ltd & Anor v Augusta/Simply , a case involving defects in the design and construction of a care home, expert accountancy and valuation evidence was required. Unfortunately while all gave helpful guidance to the court as the judge commented the Defendant’s expert was “in greater difficulties because of the complete absence of any factual evidence advanced by Simply.  Many parts of his report were based upon assumptions which were speculative and not supported by any factual evidence.” His cause was also not assisted by the lack of any mention of defects in his report. When cross examined, “he agreed he should have considered the defects but could not explain whether it was a deliberate omission or “… whether I just missed this.  I don’t know.”  This was more than somewhat surprising given the case was principally about the impact the defects had on the value and profitability of care home.”

In Ruby Properties v  James Alistair Watt  a dispute about the sale and purchase of a pilot training school the accountancy expert was not provided with data from an account system or the accounts for the company. Instead it was a document prepared in part from an accounts system and partly from witnesses recollections and assumptions and without the original records being disclosed.  Once the witness had conceded that “the spreadsheet was very much his subjective assessment which he had done from “feeling,  rather than mathematically” the accountancy experts evidence was, as the judge put it had “no factual foundation”.

It is clearly difficult for an expert when a client failed to provide the information they need.  An expert in Australia demonstrated that the best way to deal with this issue is to explain the limitations of the evidence and the limitations that places on your report. In that caser the judge concluded that within those limitations the expert “was entitled to express significant concern as to the profitability and financial viability of the entities”  and that “the evidence enables the Court to have confidence in his valuation opinion”.

The moral of these stories is to be wary of incomplete or unsupported facts and make sure that you state any issues or limitations in your report. For advice and guidance on giving expert evidence, including our  Model Form of Expert’s Report and  practical advice and guidance on topics such as GDPR, Joint Statements, privilege and so on see the Academy’s Knowledge Hub.

ure is from The Guardian.