Speed must not replace scrutiny as AI races into medicolegal

Artificial intelligence is transforming how medical experts handle the vast bundles of records at the heart of personal injury and clinical negligence claims. TLA Medicolegal is warning that the rush to adopt it must never come at the expense of analysis.

Some experts now use AI to organise, paginate and index bundles that can run to several thousand pages, most of which are medical records sitting alongside witness statements and other experts’ reports. Software can now tell an expert in moments what imaging has been carried out, when the last scan was taken and which records are missing.

Adoption is accelerating, with the Bond Solon Expert Witness Survey 2025 finding that 20% of experts had used AI in their role, double the 9.3% recorded the year before, while 89% felt specific guidance was needed on the use of AI by expert witnesses in the UK.

Fiona Morrison, Director and Co-founder of TLA Medicolegal, says: “Many people are jumping into using AI in medicolegal work really rather quickly and the technology is progressing so fast that it can be hard for anyone to keep pace.

“It is an excellent tool for putting records in order and we fully support experts who use it to work more efficiently. The concern comes when organising the notes starts to replace reading them, because it only takes one entry buried in thousands of pages to change the outcome of a case. The value of an expert has always been their judgement and that is something no software can provide.

“Experts must also be completely transparent about how AI has been used in their reports, choosing tools listed on NHS England’s ambient voice technology registry such as Heidi, Tortus, Tandem, Accurx or Microsoft Dragon and confirming that the opinion is entirely their own. Those who get this right now will be in the strongest position when the rules change.”

Regulation is now catching up with practice, after a Civil Justice Council consultation earlier this year proposed extending the expert’s statement of truth, so the expert explains any use of AI beyond transcription or other administrative tasks and identifies the tools used. The Council’s direction of travel is towards proportionate transparency for expert evidence and its final report is expected later in 2026.

Guidance from the Academy of Experts takes a similar line, treating administrative uses such as document organisation as low risk while classing AI that analyses data, generates report content or helps form an opinion as high risk, requiring a high level of human oversight and transparency.

Experts using AI should include a declaration in their reports confirming that AI was used solely for dictation and transcription, that the software complies with NHS data requirements and that the opinion reflects their own scrutiny of the records, imaging and assessment.

TLA works closely with its experts to ensure every report stands up to scrutiny, combining the efficiency that new technology offers with the thoroughness that courts and solicitors rightly expect.