Following a consultation process earlier in the year, the Civil Justice Council have published their final report on AI in preparing court documents for use in civil proceedings in England and Wales. The results are more pragmatic and limited than some of the proposals in the Consultation Paper. However, the CJC continues to recommend important changes to the declarations required on the use of AI in witness and expert evidence. This could see significant changes coming to the CPR in England and Wales that civil litigation practitioners will need to be on top of.
The key recommendations
Statements of case, skeleton arguments and other advocacy. The report rejects new AI-use declarations for statements of case, skeleton arguments and other advocacy. The fact that legal representatives have professional responsibility for the content of these documents is sufficient. The CPR should though make provision for who is responsible for the document if AI is used.
Disclosure. No new rules concerning disclosure are required, subject to recommendations of the Disclosure Working Group, with the approach to AI use being guided by the existing requirements for disclosure.
Witness evidence. The CPR should be amended so that witnesses of fact are required to declare that AI has not “generated or influenced” their statement’s substantive content. This would not apply to statements prepared and signed by legal representatives for procedural purposes.
Expert evidence. The CPR should be amended to require experts to give a new declaration concerning AI use, if it generates or influences the report. The declaration should provide sufficient information for the court and opposing party to meaningfully scrutinise the way in which AI has been used. That may include identifying the AI tool where necessary. The CJC’s Guidance for the Instruction of Experts in Civil Claims (2014) should introduce good practice guidance on AI use.
Translation. There is no need to introduce new guidance or rules. As before, all translations (including AI translations) should be certified by a professional human translator and accompanied by the original text.
Clearer guidance is a priority. The report calls for coordinated regulatory guidance and training, a plain-language Practice Direction consolidating the courts’ approach and sanctions, and consideration of guidance for litigants-in-person.
In general, the new Practice Direction should define the scope of AI use functionally. The Practice Direction should not categorise tools as “generative”, “administrative” or otherwise and should identify activities capable of influencing substantive content
Comments and reflections
Overall, the report’s approach is likely to be perceived by the market as welcome. A large number of respondents to the consultation had expressed concern that a more prescriptive approach would lead to satellite litigation and become increasingly problematic as AI is embedded into more tools that may be used by the legal profession in the ordinary course of business. For the most part, that concern seems to have been addressed.
However, particularly in the context of witness statements, further guidance will required on what amounts to an AI “generating or influencing” content. “Influencing” in particular is potentially wide-ranging and it is not hard to see how AI use that is not intended to be prohibited (e.g. preparing a list of questions for a witness) could be argued to be “influencing” content. It is hoped that the guidance in the new Practice Direction provides further clarity on that issue, so that the other form of guidance – case law generated via satellite litigation – is minimised.
