TL;DR

UK courts are tightening expectations around AI use in legal practice. Recent cases have exposed the risks of unverified AI outputs, and updated judicial guidance makes clear that solicitors bear personal responsibility for all AI-generated material submitted to court.

Courts Confront AI Hallucinations

The cases of Ayinde v London Borough of Haringey and Al-Haroun v Qatar National Bank mark one of the first times the Divisional Court has directly addressed AI misuse by legal professionals. Dame Victoria Sharp, president of the King’s Bench Division, warned of “serious implications for the administration of justice and public confidence” if AI is misused, and criticised the submission of pleadings containing fabricated citations produced by generative AI.

Sharp found it “extraordinary that the lawyer was relying on the client for the accuracy of their legal research, rather than the other way around.” The court ruled that submitting unverified AI-generated content could amount to professional misconduct and result in wasted costs orders or regulatory referrals.

Updated Guidance for Solicitors

Judicial guidance published in October expanded on earlier versions by adding AI terminology, addressing bias in training data, and spelling out the risks of AI hallucinations. The key points are direct:

  • Verification is non-negotiable. All AI-assisted research, drafting, and evidence must be checked for accuracy. Failure risks wasted costs and reputational damage.
  • Ethical duties extend to AI. SRA Principles require solicitors to understand the limitations of the tools they use.
  • Confidentiality matters. Privileged information should never be entered into public AI platforms.
  • Internal governance is expected. Firms should implement protocols for AI use, including training, oversight, and policies on consumer AI tools.

What Comes Next

With the UK’s AI Regulation Bill delayed, the legal profession is operating under soft law guidance and professional standards rather than statute. Pressure for statutory intervention continues to mount. For solicitors, the immediate priority is clear: check every AI-generated output, stay current with guidance, and build internal processes that manage the risks before they reach a courtroom.

Share this article