Skip to content
Submit incident
AII-1001-40853
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.Supporting sources are pending verification.

UK Barrister Found Using AI to Generate Fictitious Legal Cases

An immigration barrister was found by a judge to have used AI software to prepare legal research for a tribunal hearing, citing cases that were "entirely fictitious" or "wholly irrelevant." The barrister failed to verify the accuracy of the citations and reportedly tried to hide his use of AI.

16 October 2025 Β· Chowdhury Rahman Β· Reliability and hallucination
16 October 2025
AII-1001-40853
Quick facts
Category
Reliability and hallucination
Country
United Kingdom
Date
16 October 2025
Drafted by
Raidu Incident Detection

What happened

Immigration barrister Chowdhury Rahman used ChatGPT-like software to prepare his legal research for a tribunal hearing in an asylum appeal case involving two Honduran sisters.

Upper tribunal judge Mark Blundell found that 12 authorities were cited in Rahman's grounds of appeal, but some of those authorities did not exist and others did not support the propositions of law for which they were cited.

Blundell stated: 'The most obvious explanation is … that the grounds of appeal were drafted in whole or in part by generative artificial intelligence such as ChatGPT.' The judge noted that one of the cases cited had 'recently been wrongly deployed by ChatGPT in support of similar arguments.'

Reported impact

Judge Blundell found that Rahman 'failed thereafter to undertake any proper checks on the accuracy' of the AI-generated content and had 'wasted' the tribunal's time.

Blundell stated that 'not one decision supported the proposition of law set out in the grounds' and rejected Rahman's appeal, finding nothing in his submissions established an error of law.

Index analysis

The AI system generated multiple fictitious legal cases and citations that did not exist, demonstrating a hallucination failure mode that the barrister did not catch before submitting to the court.

Rahman apparently tried to hide his use of AI and misled the court about the source of his legal research by claiming he had used 'various websites.'

Governance controls

Judge Blundell said he was considering reporting Rahman to the Bar Standards Board for disciplinary action.

Classification

Category
Reliability and hallucination
Industry
Government, security and defence
Country
United Kingdom
Harm status
Actual harm
Intent
Intentional
Timing
Post-deployment
Harm types
Public interest
Failure modes
Agent misaction, Fabricated citation, Hallucination, Overreliance
Consequences
Regulatory investigation
AI systems
ChatGPT

Sources and evidence

Sources not yet published

This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.

How sources are verifiedSuggest a source for this incident