Source: Bevan Brittan

The use of generative artificial intelligence tools is growing quickly in various sectors, offering efficiency improvements for tasks such as drafting and summarising documents. However, legal cases such as R (Munir) v Secretary of State for the Home Department [2026] UKUT 81 and Anthony Malcolm Cork & Anor v Mark Smith [2026] EWHC 1199 (Ch) have illustrated risks associated with AI use in legal proceedings. In Munir, the Upper Tribunal found that some legal authorities cited by solicitors could not be verified and that these references appeared to support inaccurate legal propositions, prompting scrutiny on how this content was generated and included. In the Cork case, a law firm was criticised by the court for relying on inaccurate legislation produced by AI, with the judge describing the AI outputs as often plainly wrong or very misleading. The court accepted there was no intention to mislead but noted inadequate supervision of junior lawyers.

The Tribunal in Munir also highlighted the importance of confidentiality when using AI, warning that placing confidential information into publicly accessible AI platforms risks that information entering the public domain, potentially leading to loss of legal professional privilege. Legal privilege depends on maintaining the confidentiality of communications between lawyers and clients, and once lost, privilege is typically irretrievable. Consequently, organisations are advised to adopt clear AI policies that prohibit uploading confidential or privileged material to public AI tools, ensure secure AI environments, provide training on related risks, require human review of AI outputs, and maintain records of AI usage to prepare for possible disclosure or litigation challenges. ---