AI and Expert Evidence: Opportunities, Risks and Responsibilities
This website will offer limited functionality in this browser. We only support the recent versions of major browsers like Chrome, Firefox, Safari, and Edge.
Speaking at the Bond Solon Expert Witness Conference, Mrs Justice Joanna Smith, co-chair of the working group responsible for the CJC final report on the “Use of AI for Preparing Court Documents”, addressed how AI has the potential to improve efficiency and productivity in the process of producing expert evidence, helping experts process large volumes of information, identify patterns and communicate complex material more effectively. But its rapid development does not change the fundamentals. Experts’ duties of independence, objectivity, transparency, and professional competence remain exactly the same.
Below we summarise some of the key takeaways from the conference speech:
The CJC recommends an AI declaration for experts
The CJC final report on the use of AI for preparing court documents recommended a new AI specific declaration where an expert uses AI to generate or influence the substantive content of their report, including their methodology or substantive analysis. Where required, the declaration should provide sufficient information to enable the opposing party and court to scrutinise the expert’s methodology effectively. This may, where necessary, include identifying the AI tool used. Please see our article here on the CJC final report.
AI can be used safely, but experts need to understand its limits
AI can assist with administrative tasks, information management, and drafting support, including organising and classifying documents, creating chronologies, summarising material, research, structuring reports and checking for errors or inconsistencies. However, administrative use can cross into substance. An AI summary may miss context or nuance, a transcription may inaccurately reflect what was said, and a tool used to structure material may subtly influence the content. Experts therefore need to understand what the technology is capable of and its limitations.
The key risks go beyond hallucinations, to training, understanding, and oversight
Hallucinations, confidentiality breaches, and technical misunderstandings are important risks. Inappropriate reliance on AI risks eroding independent judgment and outsourcing the expert’s opinion. Experts must verify factual assertions, references, and conclusions, rather than taking AI outputs at face value.
At a practical level, a lack of training in the relevant AI tool, insufficient understanding of what it is capable of and inadequate oversight are also significant risks. Where teams are used to analyse material or carry out research, experts should understand what AI tools team members are using and how they have been used, even where the expert has not used those tools themselves. Strict protocols should also ensure that team members understand the importance of being clear with the expert about their use of AI.
Practical safeguards should start at the point of instruction
Experts should consider from the outset whether AI use is permitted, whether the tool is fit for purpose and secure, and whether its limitations are understood. They should keep a proper record of its use and ensure outputs are validated. Where AI influences substantive content, experts should consider biases, hidden assumptions, reproducibility and whether every conclusion can be explained and defended as their own.
To this end, Mrs Justice Joanna Smith sets out a “number of important questions that every expert should bear in mind upon receipt of instructions” (verbatim as follows):
Key takeaway
AI is a tool, not an expert. Responsibility ultimately remains with the expert witness. Transparency, verification, and professional judgment should be the guiding principles. The question is not whether experts should use AI tools, but how experts can use AI tools responsibly, safely and in a manner consistent with their obligations to the court.
For more information about the law, technology and practice of disclosure, contact Tom Whittaker or Stacie Bourton.
Want more Burges Salmon content? Add us as a preferred source on Google to your favourites list for content and news you can trust.
Update your preferred sourcesBe sure to follow us on LinkedIn and stay up to date with all the latest from Burges Salmon.
Follow us