The UK's Upper Tribunal Says a Supervisor Who Fails to Ensure a Junior Fee-Earner's Work Is Free of False Citations Is Likely to Be More Culpable Than a Lawyer Who Fails to Ensure the Same of Their Own Work

The UK's Upper Tribunal is, the decision says, a superior court of record. In a decision promulgated on November 17, 2025, three judges of its Immigration and Asylum Chamber dealt with two cases in which false or wrong citations had been put before the tribunal. The decision's title ends in brackets: "AI hallucinations; supervision; Hamid." It explains that the High Court's power to govern its own procedure includes making sure that "the lawyers interacting with the court conduct themselves according to proper professional standards," and that the Upper Tribunal has for many years exercised a Hamid jurisdiction.

The decision opens on a plain duty. Legal professionals "are obliged to ensure that legal arguments which are presented to the First-tier Tribunal or Upper Tribunal are factually and legally accurate." Parts of the decision follow, not the whole, and no one it names is named here.

A Case That Never Existed

In the first case, grounds of appeal cited a case that, as far as the panel could establish, had never been reported, under a citation that belonged to an unrelated case about equal pay. The representative responsible said that no AI large language model had been used in drafting the grounds. A later witness statement repeats that he was certain he had not used a well-known AI chatbot to create the grounds, but accepts that he could not offer a cogent explanation of how the false case came to be included. "In absence of an explanation and with how AI operates, I cannot dismiss the fact that the case was an AI creation as there is no other explanation," the statement reads. His best guess, the decision records, was that he had used the "AI Mode" of an internet search engine without meaning to.

The panel put the question to the search engine's AI itself. It says that by asking the same question in slightly different ways, "we were able to elicit various different compositions of the bench which decided that fictitious case." The representative had already reported himself to his regulators, so the tribunal made no referral of its own. It says it would have made one if he had not.

Supervision

The second case turned on who checked the work. Grounds for judicial review, with citations the tribunal found "so badly wrong," had been drafted by someone the solicitor in charge described as a "very junior caseworker." That solicitor, whose name stood next to the statement of truth on the claim form was, the tribunal says, "responsible for the accuracy of the grounds for judicial review."

"It would be easy to think that this is a case about the naïve use of generative AI, but it is not merely about that; it is principally about supervision and the obligation to ensure that the tribunal is not misled," the decision says. "It matters not how such citation errors come about." The qualified professional with conduct of the matter "is expected to ensure that such documents are checked, that errors are identified, and that only accurate documents are sent to the tribunal."

In the tribunal's judgment, a supervisor who fails to ensure that the work of a more junior fee-earner 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 of them. The supervisor also fails "to aid the development of more junior lawyers."

Among the tribunal's other concerns, it says the solicitor could not tell it who had completed and signed the claim form in his name, that the case had no entry in the firm's spreadsheet of who had worked on each case, and that he could not give an accurate answer on how many other cases the junior had worked on. "Nothing is for sure," he told the tribunal. It found that the false citations and his failure to supervise the work "in this and an unknown number of other cases" necessitated a referral to the Solicitors Regulation Authority, that its concerns about his oral evidence strengthened the case for it, and says "It will be for the SRA to decide what if any action to take thereafter."

The record is part of the check. A firm that cannot say who wrote a document will struggle to show who checked it.

Checks, Forms and Confidentiality

The tribunal does not suggest, it says, that the use of legal AI programs "by properly trained professionals" is anything other than a step forward in legal practice. It also says that "any practitioner who uses non-specialist AI to undertake research or drafting is obliged to undertake rigorous checks to ensure that any information gleaned from those sources is true and accurate."

By the time of the decision, the claim form for judicial review in the Upper Tribunal "has now been amended," it says, to require a legal representative to confirm, by a statement of truth, that any authority cited in it or in documents appended to it "(a) exists; (b) may be located using the citation provided; and (c) supports the proposition of law for which it is cited." Other forms and directions, it says, "are to be similarly amended." A representative who signs such a statement in a case where false authorities are cited "should ordinarily expect to be referred to their regulatory body." Both cases before the tribunal, it notes, concern pleadings settled before those changes were made.

Among its general conclusions, the tribunal says that uploading confidential documents into an open-source AI tool places the information on the internet in the public domain, breaching client confidentiality and waiving legal privilege, and that such conduct "might itself warrant referral to the regulatory body and should, in any event, be referred to the Information Commissioner's Office."

The new form puts a signature under every citation. Whoever signs it answers for the authorities in the bundle, whoever or whatever first wrote them down.