New guidance from The Academy of Experts
Legal professionals are finding new ways for AI tools to assist them with their work, while at the same time managing the risks their use can pose. Various professional bodies have issued guidance on the use of AI, for example, the Courts and Tribunals Judiciary, and the Law Society. The most recent guidance has been produced by The Academy of Experts to assist expert witnesses who have been instructed to provide independent expert evidence in legal proceedings: Guidance for Expert Witnesses on the use of Artificial Intelligence (AI) (the Guidance).
The Guidance considers how experts may use AI in their work and the risks of using it, and suggests safeguards and practical steps experts should take. In England and Wales and other jurisdictions, experts owe important duties to a court or tribunal to provide opinions that are independent and unbiased and that only relate to matters which lie within their expertise; therefore, the risks associated with the use of AI and Gen AI technology are particularly pertinent. The Guidance will also be of assistance to lawyers and clients who instruct an expert by identifying the issues they should keep in mind and questions they may wish to ask their expert about the expert’s use of AI tools.
Appropriate use cases and levels of risk
The Guidance identifies various purposes for which an expert may consider using AI, from proof-reading drafts and translating documents, through summarising information and analysing large volumes of data, to possibly modelling or generating substantive content for the expert’s report. It suggests that experts should clearly identify the purpose for which they intend to use AI as that will enable them to determine the lawfulness, appropriateness and risk level involved in the use, plus any safeguards they should implement if they go ahead.
The Guidance discusses various practical, legal and reputational risks arising from the use of AI and particularly Gen AI. These include the risk of hallucinations and biased outputs which could seriously undermine an expert’s credibility and their client’s case, and also place the expert in breach of their legal duties to the court or tribunal. Experts should also ensure they understand how the AI tool works and how material uploaded to the tool is treated in order to manage the risk that confidentiality and legal privilege over material could be lost. For example, a recent judgment of the Upper Tribunal of the Immigration and Asylum Chamber (UK v SS for the Home Department (AI hallucinations; supervision; Hamid) [2026] UKUT 81 (IAC)) stated that uploading confidential documents into an open-source AI tool, such as ChatGPT, placed information on the internet in the public domain, and therefore would breach client confidentiality and waive legal privilege (read our article here). We have also considered this risk in the context of expert witness evidence in our article, Legal professional privilege in the generative AI era.
The Guidance suggests that experts should assess their intended use of AI for a particular purpose by reference to three categories of risk, with different considerations and approaches for each category. First, whether the use is prohibited. This prohibition could be by law or regulation (which is evolving), because it would be a breach of the expert’s duty to the court, or due to a contractual term in their engagement letter. Secondly, uses which are high-risk, either as a result of the regulatory position or because of the nature of the task, i.e. where the expert is using the AI tool to generate substantive content for their report or to undertake material analysis on which their opinion will be based. In such cases, the Guidance advises that experts should disclose any proposed high-risk use to their instructing lawyers in advance, regardless of any contractual requirement to do so. Thirdly, low-risk use such as proof-reading or using the tool to suggest stylistic suggestions without altering the substantive content of work. The Guidance suggests that such use may not need to be disclosed to their instructing lawyers, but experts should be satisfied that their instructing lawyers and clients are unlikely to be concerned, and they should be prepared to explain their use of AI under cross-examination.
The Guidance rightly emphasises that however an expert uses AI, they remain ultimately responsible for their work and the report they produce. They must ensure that their work is accurate and that they have complied with their professional duties; this requires careful consideration of any use of AI and an appropriate level of professional oversight over any AI-generated output.
Practical considerations and safeguards
The Guidance identifies various practical considerations and safeguards that experts should implement when using AI. These include:
- Choosing the most appropriate (not simply their preferred) AI tool for the purpose and ensuring they are using the latest version with the most up-to-date safety features.
- Implementing processes to ensure adequate human oversight and carefully checking Gen AI output for inaccuracy or hallucination.
- Documenting key decisions about the use of AI in the first place and keeping a record of key steps such as how the tool was used, how the output was used and the safeguards implemented. This should also include how the expert’s team members have used AI in connection with their work.
- Liaising with their instructing lawyers regarding the use of AI (which should include, but is not limited to, complying with any requirements in their engagement letter) and considering whether they are required to or should disclose their use of AI to the court and/or to the opposing side.
The Guidance also includes a helpful checklist for experts to refer to before and while using AI tools.
In summary, this is a useful addition to the growing body of guidance for the legal sector on the use of AI, for both experts and instructing lawyers. While the Guidance does not have any formal status under the Civil Procedure Rules, it may well be taken into consideration by the judiciary in relation to any disputes arising in court proceedings in relation to the use of AI in expert evidence.
Subscribe to our Inside Disputes Blog to receive regular expert insights into banking and commercial litigation in the English courts and issues relevant to investigations.