Core Event
In a filing issued on Wednesday, the New Mexico Supreme Court said attorney Stephen Aarons was fined $5,000 and held in contempt after submitting an appeal brief containing AI-fabricated testimony in a murder-conviction case. The court found that he failed to “verify the factual claims and legal authority in his AI-generated brief.”
- Authority issuing the penalty: New Mexico Supreme Court
- Person penalized: Attorney Stephen Aarons
- Fine: $5,000
- Court finding: Failure to verify factual claims and legal authority in an AI-generated brief
- Material at issue: Fabricated witnesses, false police testimony, and false testimony about the shooter’s clothing and appearance
What Happened
According to Reuters, Aarons admitted during an August hearing that he had used ChatGPT and believed it would produce a “bulletproof summary” of the trial.
The court filing said the brief “contained false testimony from wholly fabricated witnesses,” as well as false testimony concerning the shooter’s clothing and appearance. During the hearing, Justice C. Shannon Bacon asked Aarons whether he watched the news, listened to the radio, or read about current events, adding that lawyers’ reliance on AI hallucinations was “an above-the-fold story every single day.”
The case is not isolated. The source report notes that two law firms were rebuked last year for filing a brief containing “numerous false, inaccurate, and misleading legal citations and quotations.” In another case, lawyers were fined for submitting AI-generated misquotes and fake citations.
In a statement to Reuters, Aarons said he was “remorseful” but hoped the disciplinary board would recognize that it was an “honest mistake.”
A Renewed Line on Legal Ethics
The case underscores a basic limitation of generative AI in legal work: it may assist with drafting and organizing information, but it cannot replace a lawyer’s final review of material submitted to a court. Witness accounts, factual assertions, case law, and legal citations all require source-level verification.
The central risk is not merely that AI can make mistakes. It is that plausible, polished output may be mistaken for reliable legal analysis. In court filings, even a single fabricated fact or citation can affect proceedings, clients’ interests, and a court’s assessment of professional conduct.
Practical Takeaways
- For lawyers already using AI: Adopt a “generate, review, cross-check” workflow, with particular scrutiny for testimony, factual assertions, and legal citations.
- For teams handling appeals or high-stakes cases: Do not place AI-generated material directly into court filings before a reliable review process is in place.
- Key reminder: Labeling text as “AI-assisted” does not remove the filer’s responsibility for its truthfulness and accuracy.
Final Note
Legal AI is useful as an efficiency tool, not as a substitute for professional judgment. This penalty does not prohibit AI use in legal work; it reinforces that humans remain responsible for checking facts and legal claims that may affect judicial proceedings.