The age of automation: When AI errors become legal risks
As the age of AI continues to transform the economy and wider society, for legal professionals this technological revolution has brought both promise and peril. Whilst it undoubtedly presents opportunities for lawyers, a recent judgment of the High Court serves as a cautionary tale of the dangers posed by AI for legal practitioners– Ayinde v Haringey, and Hamad Al-Haroun v (1) Qatar National Bank QPSC & (2) QNB Capital LLC, [2025] EWHC 1383 (Admin).
Overview
The two cases were listed together under the so-called Hamid jurisdiction, in which the English Court has an “inherent power to regulate its own procedures and to enforce duties that lawyers owe to the court”.
Ayinde v Haringey
Here, the underlying proceedings concerned a judicial review brought against a local authority. Relevant to the Hamid jurisdiction was that one of the barristers had (i) included in their submissions citations to cases which did not exist; and (ii) incorrectly summarised a legislative provision. When the barrister’s instructing solicitor asked for copies of the cases allegedly cited, the barrister was unable to provide them. Further, in response to concerns raised about citations to non-existent cases, the barrister had drafted a response which sought to downplay the issue.
Although there was no finding that it was the use of AI specifically that had resulted in the citations to non-existent cases and the inaccurate description of legislation, the Court concluded that the barrister “could have checked the cases she cited by searching the National Archives’ caselaw website or by going to the law library of her Inn of Court. We regret to say that she has not provided to the court a coherent explanation for what happened.”
Al-Haroun v (1) Qatar National Bank & (2) QNB Capital
The underlying dispute in this case related to a claim for damages arising from the alleged breach of a financing agreement. The relevant facts for the purposes of the Hamid jurisdiction were that the solicitor had relied on his lay client’s legal research without checking it. It transpired that the ‘legal research’ was based on case citations which had been generated partly using AI. Both the lay client and the solicitor issued an apology to the Court.
Key takeaways
There are a number of important takeaways for practitioners. Although no contempt of court proceedings were initiated in either case (despite a finding in the Ayinde case that the threshold for contempt had been met), the cases highlight how a cavalier approach to using AI can give rise to conduct issues; indeed, in both cases referrals to the relevant professional regulator were made. Moreover, the Court noted that a copy of its judgment would be sent to the Bar Council and the Law Society so that they can “consider as a matter of urgency what further steps they should now take in the light of this judgment”.
Whilst this case concerns the use of AI in the dispute resolution context specifically, there is a growing trend for lawyers to use AI-assisted tools in all areas of legal practice. It bears noting that the professional conduct obligations to which lawyers are subject apply generally, whether or not lawyers are interacting with courts or tribunals.
At a time when developments in AI are front and centre of the legal industry, the judgment serves as an important reminder of the potential pitfalls of using AI for legal practitioners. Whilst the cost-saving benefits of AI are well-known, the judgment is a salutary warning that what may seem like a time-saving shortcut must be approached with great caution. For lawyers, the message is clear: professionalism and integrity can never be short-circuited.
