Fact or Fiction: The Impact of AI on Inaccurate Court Submissions

Background

UK and R (on the application of Munir) v Secretary of State for the Home Department (AI hallucinations; supervision; Hamid) [2026] UKUT 00081 (IAC) concerned two cases where legal representatives submitted false cases and citations to the Upper Tribunal Immigration and Asylum Chamber (“the Tribunal”). Whilst the primary focus of the judgement is on the submission of inaccurate legal authorities, the Tribunal made boarder observations on the impact of Artificial Intelligence (“AI”) on client confidentiality and legal professional privilege.

Use or Suspected Use of AI

The first case concerned a citation which was submitted in grounds of appeal to the Tribunal called “Horleston v SSHD [2007] EWCA Civ 654”. This case does not exist, and the citation is in fact for an unrelated equal pay case. The firm who submitted the grounds, TMF Immigration Lawyers, were asked to identify who was responsible for the case and the drafting, to provide an explanation for why a case was cited could not be located and to state explicitly whether an AI tool such as ChatGPT was used to draft the grounds. The principal of TMF Immigration Lawyers explained that he was the person responsible for the case and drafting the grounds. Whilst he initially confirmed that the reference to the Horleston case “arose as a result of human error”, he went on to concede that he “cannot dismiss the fact that the case was an AI creation as there is no other explanation”.

The second case concerned inaccuracies in grounds of judicial review submitted by City Law Practice Solicitors and Advocates (“CLP”). The Upper Tribunal Judge (“UTJ”) Blundell ordered that CLP identify the author of the grounds, explaining that several authorities could not be found or were incorrectly used. One particular citation was one of UTJ Blundell’s own cases, which was incorrectly cited as a High Court case and the argument extracted from the case was incorrect. CLP’s Compliance Officer for Legal Practice (“COLP”) identified that a “part-time trainee lawyer” had drafted the grounds under his supervision. The COLP apologised for the inaccurate citations and stressed that there was no intention to mislead. However, the Tribunal held that a supervisor who fails to ensure that the work of a junior does not contain false cases or citations is “likely to be more culpable than a lawyer who fails to ensure that his own work is free from such “hallucinations””.

Client Confidentiality and Legal Privilege

In the first case of Munir, the firm was asked to provide the Court with examples of how TMF Immigration Lawyers use AI in order to prevent similar occurrences. They explained that they had used ChatGPT to summarise Home Office decision letters and to suggest improvements to client letters. The Court observed that putting Home Office decision letters and client letters into an open source AI tool, such as ChatGPT, places the information in the public domain and therefore breaches client confidentiality and waives legal privilege.

Conclusion

Munir considers the consequences of submitting false information before the Court. The Court warns legal professionals of their duty to ensure the Courts are not misled, highlighting the wasted time associated with verifying the information’s accuracy and the potential loss of public confidence in the processes of the Court. The Court’s powers include public admonition of the lawyer, the imposition of a costs order, the imposition of a wasted costs order, striking out a case, referral to a regulator, the initiation of contempt proceedings, and referral to the police. In Munir, both cases concluded that referrals to the Solicitors Regulation Authority were necessary.

Unfortunately, the Tribunal in Munir highlighted that it has seen “a considerable increase in the latter half of 2025 in the citation of fictitious authorities”. Both cases highlight the issue of a lack of understanding as to the uses and limitations of AI. The Law Society of England and Wales and the Law Society of Northern Ireland have both issued guidance for solicitors in the use of AI within the profession. More recently, the SRA has published a warning notice on “Misuse of AI”, identifying AI hallucinations and client confidentiality as two particular areas of concern. As the prevalence of AI increases, it is particularly important for legal professionals to familiarise themselves with the guidance available.

Key Takeaways for Legal Professionals

i. In Munir, uploading confidential information into open-source AI tools was considered to be placing the information in the public domain and therefore it breaches client confidentiality and waives legal privilege.

ii. There is a greater burden on supervisors who fail to ensure the work of their junior fee-earners do not contain false cases or citations than the fee-earner who fails to ensure that their own work does not contain false cases or citations.

iii. Legal professionals should ensure that, if they are using AI tools in legal research, they are undertaking rigorous checks to satisfy themselves that the information is true and accurate.

If you would like any further information or advice on this matter, please contact a member of the Professional Negligence Team.

*This information is for guidance purposes only and does not constitute, nor should be regarded, as a substitute for taking legal advice that is tailored to your circumstance.