News & Insights
James Lloyd discusses the judgment handed down on 6 June 2025 in Ayinde v London Borough of Haringey and Al-Haroun v Qatar National Bank [2025] EWHC 1383 (Admin), in which the Divisional Court issued a powerful warning to legal professionals about the risks of using generative artificial intelligence tools for legal research and drafting, and provided a stark reminder of the severe consequences for both lawyers and litigants who knowingly or inadvertently present ‘fake’ material to the court.
The combined cases of Ayinde v London Borough of Haringey and Al-Haroun v Qatar National Bank [2025] EWHC 1383 (Admin) involved the use or suspected use of generative AI resulting in fictitious caselaw, fake citations, and misstatements of law being relied upon by parties to litigation. The cases were heard under the inherent jurisdiction of the High Court to regulate its own procedures and enforce lawyers’ duties to the court (R (Hamid) v SSHD [2012] EWHC 3070 (Admin)).
Ayinde – pupil barrister allegedly using AI
The facts of Ayinde will be well-known to most lawyers, given recent publications following the wasted costs application on 3 April 2025. In the course of judicial review proceedings relating to interim accommodation, counsel instructed on behalf of the claimant relied upon case law in correspondence of which the Court said:
The case that is cited (El Gendi) does not exist. There is no case with that name, held by the National Archives, or anywhere else. The neutral citation number, [2020] EWHC 2435 (Admin), does exist, but it is the citation reference to a different case. [36]
In the settled grounds, further concerns arose. The Court noted that the settled grounds significantly misstated the effect of s.188(3) of the Housing Act 1996 and:
… four further cases cited by Ms Forey do not exist either. We note too the Americanised spelling of “emphasized”, which contrasts with the English spelling of the same word by Ms Forey in correspondence; and further, the somewhat formulaic style of the prose. [38]
The solicitor for the defendant raised concerns about the citations and misstatement of law in pre-hearing correspondence. Those concerns did not prompt, in the Court’s judgment, sufficient scrutiny by the claimant’s solicitor. Rather, the claimant responded dismissing the concerns as “easily explained” and suggesting that the defendant lawyers “may better serve your organisation by giving attention not to the normative discoveries you have made, but whether you can locate the authorities in support of the points raised.” [42]
Before the Divisional Court, as in the earlier wasted costs application, the suggestion that counsel had relied upon Chat GPT or some other LLM was disputed. An explanation, rejected by the Divisional Court, was given. Within counsel’s submissions, criticism appears to have been levelled at the chambers at which counsel was, at the relevant time, a pupil barrister:
[Counsel] says that she had received little formal supervision during her first six months pupillage. She does not recall attending court with a member of chambers in relation to a claim for judicial review in a homelessness case. She says that during her second six months pupillage she had an extremely busy practice in her own right. She did not receive any supervision. None of her written work was checked. [51]The tenor of those submissions, the veracity of which was not determined, garnered little sympathy:
We do not accept that a lack of access to textbooks or electronic subscription services within chambers, if that is the position, provides anything more than marginal mitigation. Ms Forey 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. [67]
Ultimately, the Court concluded that:
On the material before us, there seem to be two possible scenarios. One is that Ms Forey deliberately included fake citations in her written work. That would be a clear contempt of court. The other is that she did use generative artificial intelligence tools to produce her list of cases and/or to draft parts of the grounds of claim. In that event, her denial (in a witness statement supported by a statement of truth) is untruthful. Again, that would amount to a contempt. In all the circumstances, we consider that the threshold for initiating contempt proceedings is met. [68]
That the Law Centre acting for the claimant “relies heavily on the expertise of specialist counsel” and that it had “not been its practice to verify the accuracy of case citations or to check the genuineness of authorities relied on by counsel” [59] was not a satisfactory retort in the Court’s judgment:
We accept that they are an overstretched charity providing an important service to vulnerable members of society with limited resources. It could be said however, that in those circumstances, it is all the more important that professional standards are maintained, and they instruct those who adhere to them. [71]
Al-Haroun – lay person using Chat GPT; lawyers following blindly
Al-Haroun sought damages for alleged breaches of a financing agreement. His solicitor was Mr Hussain of Primus Solicitors. In support of an application to set aside an order, witness statements in the names of Al-Haroun and Mr Hussain contained
… numerous authorities, many of which appear to be either completely fictitious or which, if they exist at all, do not contain the passages supposedly quoted from them, or do not support the propositions for which they are cited. [73]
Al-Haroun “accepts responsibility for the inclusion of inaccurate and fictitious material in the witness statement that he filed with the court. He says that the citations were generated using publicly available artificial intelligence tools, legal search engines and online sources. He had complete (but he accepts misplaced) confidence in the authenticity of the material that he put before the court.” [76]
The solicitor accepted that he “relied on legal research that Mr Al-Haroun had conducted, without independently verifying the authorities”. [77]
The Court was scathing in its criticisms of the solicitors involved:
it is extraordinary that the lawyer was relying on the client for the accuracy of their legal research, rather than the other way around [79]
As to Mr Hussain, and Primus Solicitors, there was a lamentable failure to comply with the basic requirement to check the accuracy of material that is put before the court. A lawyer is not entitled to rely on their lay client for the accuracy of citations of authority or quotations that are contained in documents put before the court by the lawyer. It is the lawyer’s professional responsibility to ensure the accuracy of such material. [81]
The Court’s response – referral not contempt
In both cases, the Court declined to initiate contempt proceedings.
As to Ayinde’s counsel, the Court decided that the factual issues “could not easily be determined in the course of summary proceedings for contempt,” instead referring the matter to the Bar Standards Board for investigation to include:
questions raised as to potential failings on the part of those who had responsibility for training Ms Forey, for supervising her, for “signing off” her pupillage, for allocating work to her, and for marketing her services. Those could not be addressed in contempt proceedings brought against Ms Forey alone. [69]
As to Ayinde’s solicitor, the Court concluded that there was no basis to suspect that he had deliberately caused false material to be put before the court, but referred the matter to the Solicitors Regulation Authority to investigate:
the steps taken by [the claimant’s solicitor] in response to the correspondence from [the defendant]; and the steps he took to satisfy himself that Ms Forey had sufficient experience or was competent to undertake the work she had been instructed by Haringey Law Centre to do. [72]
The Court accepted that Al-Haroun had not intended to mislead the Court, and that he had misplaced trust in online resources.
As to Mr Hussain, and Primus Solicitors, the Court referred the matter to the Solicitors Regulation Authority to investigate.
The outcome of (any) future regulatory proceedings remains to be seen. It can be expected, however, that regulators shall take a dim view of any evidenced reliance upon ‘fake’ caselaw, particularly given the Court’s concerns in this case and the frequency with which these issues now arise. There appears to be an emerging trend which regulators might rightly wish to arrest. As recently as May this year in Bandla v Solicitors Regulation Authority [2025] EWHC 1167 (Admin), for example, the Court encountered the same phenomenon:
In my judgment, the Court needs to take decisive action to protect the integrity of its processes against any citation of fake authority. There have been multiple examples of fake authorities cited by the Appellant to the Court, in these proceedings. They are non-existent cases. Bandla at [53]
What action must now be taken by practitioners?
All lawyers should be familiar with the rules engaged by the use of LLMs.
For solicitors, the SRA Code of Conduct articulates clear obligations:
For barristers, the BSB Handbook [v4.8] is clear:
Despite the wealth of supplementary guidance and regulator commentary available to lawyers, underpinning the very clear duties in our respective codes of conduct, the Court’s message is clear:
More needs to be done to ensure that the guidance [on AI use] is followed and lawyers comply with their duties to the court. [82]
Though the Court arranged for the judgment to be sent directly to the Law Society and Bar Council, the Court’s expectations do not end with the regulators. Firms and Chambers can now expect a greater level of scrutiny where instances of AI misuse arise:
There are serious implications for the administration of justice and public confidence in the justice system if artificial intelligence is misused. In those circumstances, practical and effective measures must now be taken by those within the legal profession with individual leadership responsibilities (such as heads of chambers and managing partners) and by those with the responsibility for regulating the provision of legal services. Those measures must ensure that every individual currently providing legal services within this jurisdiction (whenever and wherever they were qualified to do so) understands and complies with their professional and ethical obligations and their duties to the court if using artificial intelligence. For the future, in Hamid hearings such as these, the profession can expect the court to inquire whether those leadership responsibilities have been fulfilled. [9]
Hence, practitioners must now act.
Individuals
Any individual practitioner who chooses to use AI tools to supplement research or drafting must do so with redoubled caution, understanding that these tools remain flawed, despite their extraordinary utility, and cannot be relied upon for accuracy.
It shall be no defence to allegations of misconduct that deference was paid to an AI tool.
Firms
The Court’s concerns in Ayinde that the solicitors may not have sufficiently considered the competence and experience of counsel before providing instructions, is a stark reminder. Firms must ensure that they have in place measures to assure themselves that counsel instructed in any given case is sufficiently experienced.
Junior and senior solicitors alike should be reminded of the Court’s scathing criticisms in this case: where caselaw is cited in pleadings or correspondence, even where drafted by counsel, checks as to the accuracy of the citation and quotation should always be made.
Managing partners and leadership teams should take heed of the Court’s warning that future cases of AI misuse shall prompt inquiry by the Court as to sufficiency of training received by offending lawyers.
Chambers
Chambers must ensure that the “leadership responsibilities” outlined by the Court in this case are discharged, reviewing their formal training on AI use and drafting policies and internal guidance.
As the profession continues to realise and embrace the enormous potential in the adoption of AI-augmented working, we must do so cognisant of its dangers. Though many examples have reached us from other jurisdictions over recent years, Ayinde provides a timely and local reminder that responsibility for the accuracy and integrity of work cannot be delegated to technology. Technologies such as ChatGPT and CoPilot are simply tools. They do not replace the craftsman, nor will the courts allow a bad workman to blame them.
James Lloyd specialises in regulatory matters, professional discipline, and emerging issues in AI governance and misuse. He has acted in recent cases involving the improper use of AI in regulatory contexts, including advising on professional conduct proceedings arising from AI-generated material and AI-generated material being used to facilitate fraud.
James shall be speaking on a panel discussing emerging AI trends in the fraud, insolvency and asset recovery sectors across different jurisdictions at TL4FIRE’s Middle East conference in Dubai on 9 November. James’ speaker code FIREMESPK25 may be used for a 10% discount.
Tom Edwards looks at the impact of the shift from Joint Enterprise to Common Purpose in the five years since…
Ben Hargreaves explores the inherent challenges in the admissibility of sexual history in sex cases. Section 41 of the Youth…
Silas Lee, pupil barrister, reviews the statutory regime on witness anonymity. Anonymous witness orders are most commonly sought by the…
An analysis of the law on fitness to plead and stand trial in the magistrates’ courts: Silas Lee reviews the…
