Now available: The Expert Record, the Standard for expert witnesses using AI. Where to buy

The Expert RecordThe standard for expert witnesses using AI

AI in the Courtroom · Expert witnesses and AI

LR (Male VOTs) Albania CG [2026] UKUT 00315 (IAC)

In LR (Male VOTs), a country guidance case, the Home Office's advocate suggested in cross-examination that the appellant's expert had written his report with generative AI. The Upper Tribunal found no proper foundation for the allegation on August 13, 2026. It reduced the weight of the report for a different reason: the expert had overlooked a recent source.

Primary document readAllegation of AI authorship rejected as lacking any proper foundation; report's weight reduced on a separate ground August 13, 2026

What happened

LR (Male VOTs) is a country guidance decision of the Upper Tribunal, reported as [2026] UKUT 00315 (IAC). Country guidance cases set the factual baseline that immigration judges across the United Kingdom apply to later appeals from the same country, so the expert evidence in them is tested at length. This one concerned male victims of trafficking from Albania.

One of the appellant's experts was a university professor with long research experience in Albania. The Home Office put 199 written questions to him, and his answers ran to a further 69 pages. He then gave oral evidence for a full day of the hearing.

In cross-examination, the Home Office's advocate suggested that the professor's report was not his own work and had been generated, in whole or in part, by generative AI. The decision records that the advocate developed the point at length. He relied on three things: the same phrases appeared repeatedly in the report, the report did not mention a 2025 trafficking report that was the most recent available, and the professor had written "to my opinion."

What the tribunal did

The tribunal rejected the allegation in two paragraphs. It did not consider there to have been any proper foundation for what it called a serious allegation, and it said none of the three points even began to support it.

The panel took the points one at a time. The repetition had an explanation: the professor said the questions from the appellant's solicitors overlapped, so his answers did too. The other two points, the tribunal reasoned, pointed away from AI. A system would have been more likely to cite up-to-date material and to use correct English. The professor explained that he had overlooked the newer report, and that English is not his first language. The tribunal doubted that the phrasing question should have been put at all to an academic writing in his second language, and it noted that he used the same expression in his oral evidence.

The oversight still had a consequence. The LR (Male VOTs) decision states that the failure to consider the most recent report would affect the weight given to the professor's evidence. The tribunal separated that ordinary criticism from the charge of AI authorship. It also rejected the Home Office's further suggestions that he was engaged in advocacy or lacked impartiality.

Where it stands

LR (Male VOTs) was promulgated on August 13, 2026 and is a reported decision. No appeal was located as of November 16, 2026. The paragraphs on the AI allegation are a small part of a long decision; the country guidance itself is outside the scope of this record.

The obligation it engaged

LR (Male VOTs) belongs in this record because the accusation is becoming a cross-examination tactic, and this tribunal said what it requires. Style is not evidence. Repetition, an awkward phrase and a missed source did not amount to a foundation, and the tribunal treated the charge as serious enough that it should not be made without one. The decision sits beside the Bard MDL ruling in the United States, where suspicion of AI use was also held to be insufficient.

For the expert, the matter shows the value of Clause 7 of the Standard, contemporaneous record, from the other side. A record of each system used and what it was used for also establishes what was not used. An expert who keeps one can answer an accusation of AI authorship with a document, and an expert who used no system at all can say so and show the drafts. The professor in LR (Male VOTs) had to rely on his own explanation and on the tribunal's reading of his prose.

Clause 8, disclosure, is the related clause. An expert who states at the outset whether any system was used removes the question from cross-examination before it is asked.

The decision also carries a caution. The allegation failed, but the missed source cost the expert weight. Completeness of sources is tested whether or not AI is involved, and Clause 3, independent authority, is of no help to a report that omits the current material.

Sources

What could not be confirmed

Paragraphs 153 through 160 of the decision were read in full on the GOV.UK tribunal decisions site; the remainder of a decision of roughly 65,000 words was not read. The professor and the advocate are named in the decision; this record does not name them because the allegation was rejected. The statement that such accusations are becoming a cross-examination tactic is this record's observation and not a finding of the tribunal. Any appeal was not located as of November 16, 2026.

About this entry

This entry was checked against the primary document listed in the sources. The entry for LR (Male VOTs) Albania CG [2026] UKUT 00315 records conduct and what the tribunal did. It does not characterize a holding and it is not legal advice. Last checked November 16, 2026. More matters like this one are collected under AI use by expert witnesses. If something here is out of date, write to [email protected].

Top