On 29 October 2026, the Civil Justice Council (CJC) published its final report on the use of AI for preparing documents provided to a court in civil proceedings. The final report followed the CJC’s interim report and consultation process and focuses on the use of AI by legal representatives and expert witnesses. The report makes 15 recommendations. Overall, the CJC took an incremental approach. It confirmed that the use of AI in court documents by legal representatives is a matter for professional regulation but proposed this should be supplemented by a new Practice Direction to codify the court’s existing approach to the use, and sanctions for the misuse, of AI by legal representatives and provide more guidance. Interestingly, it recommended new procedural requirements in relation to witness statements and expert evidence, which would be implemented through new rules prepared by the Civil Procedure Rule Committee. 

More limited recommendations than interim report

The final recommendations are more limited than some of the proposals in the interim report. 

For example, the CJC concluded that no new rules should be introduced requiring legal representatives to make a declaration concerning AI use in the preparation of statements of case, skeleton arguments or other forms of advocacy. Existing professional duties on lawyers protect the integrity of such documents, provided that the individual lawyer who takes responsibility for the document is identified. The CJC also considered that no procedural reforms are currently needed regarding AI use in the disclosure process (although it noted that further rules may be suggested in future following the separate review of PD57AD by the Disclosure Review Working Group). Similarly, the CJC determined that there is no need to introduce any new substantive rules concerning the use of AI translation tools – all translations should continue to be certified as accurate by a professional human translator.

Recommendations for witness statements and expert evidence

The two most significant recommendations relate to the preparation of witness statements and expert evidence; the CJC proposed amendments to the CPR to require specific declarations regarding AI use. 

A witness of fact who makes a witness statement (whether for trial or otherwise) should make a declaration that AI used in its preparation has not generated or influenced its substantive content. (The declaration would not be required for witness statements prepared and signed by legal representatives for procedural purposes as these are subject to existing professional duties.) The CJC expressed concern that the use of AI that can generate or influence substantive content poses a significant risk to the integrity of evidence, including the essential requirement that witness statements of fact must record the witness’s evidence in their own words. The proposal does not prohibit the use of any AI tool in the preparation of witness statements, but focuses on the result of AI use on the substance of the evidence. 

The position of experts is different as they may use AI in their methodology or substantive analysis. However, to maintain the integrity of expert evidence, an expert should declare AI use where such use has generated or influenced the substantive content of the expert’s report. The declaration should provide sufficient information (including, where necessary for this purpose, disclosure of the AI tool used) to the opposing party and the court, to facilitate effective scrutiny of the expert evidence, including the methodology used, the appropriateness of AI use for that purpose, and its substantive content. The CJC acknowledged that the existing obligations on experts to disclose their methodology may be understood to already require such disclosure but considered it preferable for there to be a separate and distinct obligation regarding AI use. The CJC also recommended revisions to the CJC’s Guidance for the Instruction of Experts in Civil Claims (2014) to include good practice guidance on AI use by experts.

The CJC rejected a technical definition of AI or categorisation of tools as generative or administrative, instead recommending that the CPR should define the scope of AI use functionally. The report therefore focuses on AI use that generates substantive content (i.e. where an AI tool produces or modifies substantive content whether intentionally or unintentionally) or influences substantive content (i.e. where an AI tool without itself producing such content, affects its selection, analysis, development or evaluation). 

In practice, the difference between form and substance in factual and expert evidence is often not entirely clear and a matter of debate. We will need to await the draft new rules to see how this potential ambiguity is dealt with. However, inevitably when preparing evidence and making such declarations, careful thought will need to be given to where the line falls between AI helping with that preparation and AI impacting the substance of the evidence. The report notes that a basic spell check or AI transcription which reproduces the source material accurately does not generate substantive content. However, beyond that, solicitors and clients will need to consider at what point the use of AI tools to suggest more substantive edits, or to re-order or organise the transcribed material crosses the line. If the recommendations are implemented, it will become even more important for solicitors, experts and clients to keep clear records of how they have used AI in the production of evidence and of discussions with witnesses about their evidence, and ensure that the use of AI is addressed in the instructions to any expert witness.