A Maryland lawyer has been sanctioned by a federal appeals court after admitting to using a generative AI tool to draft a brief that cited multiple nonexistent cases.
Kusmin L. Amarsingh on Monday was sanctioned $1,000 by the US Court of Appeals for the Tenth Circuit, which also affirmed the dismissal of her racial bias and other claims against Frontier Airlines Inc.
In addition to the monetary sanction, the Tenth Circuit referred a copy of its order and judgment to attorney-disciplinary authorities in Maryland, where she’s licensed.
Amarsingh is the latest of a growing list of lawyers being sanctioned by federal judges after admitting they misused generative AI tools and submitting “hallucinated” mistakes in court filings without catching and fixing them first. Judges themselves and their clerks are also getting caught up in the mix.
The Tenth Circuit’s review of Amarsingh’s brief revealed “serious flaws” in her appellate brief, most prominently that she had relied on ChatGPT before citing “multiple nonexistent cases” and misattributed propositions and quotations to actual cases, said Judge Harris L. Hartz in the unpublished opinion.
After finding the mistakes, the appeals court ordered her to explain and correct the fiing. She admitted that she had used ChatGPT, which generated the seven fabricated case citations, to help research and drafting the brief..
Amarsingh said she had no intention of misrepresenting the law or the record, the opinion noted.
The Tenth Circuit sanctioned Amarsingh under Federal Rule of Appellate Procedure 38. The test under Rule 38 isn’t whether an attorney acted with subjective bad faith but whether the conduct, viewed objectively, “‘manifests either intentional or reckless disregard of the attorney’s duties to the court,’” the appeals ourt said.
“Amarsingh’s actions were reckless because of her complete failure to perform an attorney’s fundamental duty to the court—confirming that all the cited cases exist and stand for the propositions for which they are cited, and accurately quoting from the cited authorities,” Hartz wrote.
The attorney said she is committed to preventing similar future errors, and has already completed a continuing legal education course on the ethics of using artificial intelligence, the opinion also noted, part of a growing CLE trend.
“We appreciate Amarsingh’s candor regarding both her use of GenAI and her failure to verify the results. We also appreciate that she is willing to try to do better in the future,” the Tenth Circuit said.
After a series of verbal altercations with Frontier gate agents during a trip from Philadelphia to St. Louis, Amarsingh alleged breach of contract and violations of federal civil rights law, and sought $15 million in damages. She alleged the gate agents discriminated against her because she isn’t Black. The lower court granted Frontier’s motion to dismiss.
The Tenth Circuit agreed with the district court that Amarsingh did not plausibly allege a federal civil rights claim.
Amarsingh contended that Frontier has a history of hostility that amounts to discrimination. “But she points to no specific episodes, relying on general complaints in the news and online,” the appeals court said.
The Tenth Circuit court also rejected her argument that the district court “disregarded essential evidence,” saying she failed to identify “what evidence the district court disregarded or what material facts it failed to address.”
Judges Nancy L. Moritz and Veronica S. Rossman joined the opinion.
Amarsingh represented herself. Fitzpatrick, Hunt & Pagano LLP, and Werge & Corbin LLC represent Frontier.
The case is Amarsingh v. Frontier Airlines Inc., 10th Cir., No. 24-01391, unpublished 2/9/26.
To contact the reporter on this story:
To contact the editor responsible for this story: