US judge sanctions four attorneys after AI-generated bogus citations found in briefs

A federal judge in Mississippi removed all four lawyers from an active lawsuit and imposed a combined $8,000 in fines on Monday after attorneys on both sides of the case admitted to submitting court filings containing fictitious citations produced by AI tools.

Business Insider reported on the sanctions order, in which US District Judge Sharion Aycock also barred two of the four lawyers from appearing before the court for two years.

What the judge found

The 23-page sanctions order was issued in connection with a contractual dispute over legal fees between plaintiff and Louisiana Attorney Tom Withers and the defendant, the Mississippi city of Aberdeen.

In the order, Aycock wrote that the US District Court for the Northern District of Mississippi "is yet again 'burdened with addressing AI hallucinations in court filings,'" adding: "This case presents the Court with an unusual scenario — attorneys for both litigants engaged in similar sanctionable conduct."

Kathleen Wilson, an out-of-state attorney representing Withers, admitted to using an AI tool to conduct legal research. Kathryn Williams, an out-of-state lawyer acting for Aberdeen, acknowledged using generative AI to draft a filing. 

"Neither of them verified the legal authority output by AI before filing their briefs," Aycock wrote. "Their acts of relying on AI output without verification alone supports a finding that they acted in bad faith."

Shauncey Ridgeway and Mark McClinton, the local counsel for the plaintiff and defendant respectively, admitted they had not reviewed the filings before submission. All four lawyers said they only became aware that AI-hallucinated cases had been cited after the judge flagged the issue late last year. They apologised and, in Aycock's words, "expressed embarrassment" — though the apologies did little to satisfy the court.

Penalties handed down

Wilson and Williams received the heaviest sanctions — two-year suspensions from the court and fines of $2,500 and $3,500 respectively. Ridgeway and McClinton were disqualified from the case and each fined $1,000.

The discovery of the errors had previously prompted the court to stay the case entirely and cancel the scheduled trial. Withers and the city of Aberdeen have been given 60 days to find new legal representation.

Aycock also weighed in on the risks of rubber-stamping AI-generated work. "In an era of rampant unverified AI usage within the legal field, this case presents a prime example of the risk associated with serving as a rubberstamp when acting as local counsel," she wrote.

Williams declined to comment, while the other three attorneys did not respond to requests for comment before publication. A Managing Partner at Ridgeway's firm, Christian & Small LLP, said the practice would "continue to educate our team about the appropriate use of artificial intelligence tools when they can benefit our clients, and the absolute requirement that our lawyers verify all information in our filings is accurate and correct."

A pattern of AI errors in courts

The Mississippi case is part of a wider trend that has drawn increasing judicial frustration across the United States. Mark Bartholomew, a Professor at the University at Buffalo School of Law, told Business Insider that early offenders were often let off lightly, but that posture has shifted.

"In some earlier examples of the same behavior, offending attorneys were let off with a slap on the wrist," he said. "Now judges just don't buy it when a lawyer protests they had no idea that AI models can hallucinate, and they are willing to call such behavior out as bad faith."

The legal profession has embraced AI at a rapid pace — the 2026 Legal Industry Report by business platform 8am found that 69 per cent of surveyed legal professionals reported using generative AI tools for work. High-profile examples of AI-related errors have grown alongside that adoption.

Last year, law firms Ellis George and K&L Gates were ordered to pay approximately $31,000 in sanctions after attorneys filed a brief containing AI-generated fabricated cases. Earlier this year, a partner at Wall Street firm Sullivan & Cromwell apologised to a federal bankruptcy judge after submitting a filing containing AI hallucinations.

Bartholomew warned that more sanctions are likely to follow. "The temptation is just too great for overworked lawyers looking for ways to cut corners," he said, adding that until AI companies "solve the hallucination problem, there will be more of these cases and more sanctions."