On Wednesday the New Mexico supreme court handed down a $5,000 fine and a contempt finding against defence attorney Stephen Aarons after he submitted a legal brief that contained wholly fabricated police testimony and invented witness statements produced by ChatGPT. The judgment marks another instance where a lawyer has been penalised for relying on AI output without proper verification, highlighting the growing tension between cutting‑edge technology and courtroom integrity.
The AI‑Generated Brief and Its Fabrications
Aarons told the court he had asked ChatGPT to create a “bulletproof summary” of the trial transcript when he took over the appeal of Oscar Renee Sandoval last year. Instead of a concise recap, the AI produced a passage that included false testimony from witnesses that never existed, such as a claim that the shooter was wearing dark pants and a white shirt. The justices described the filing as containing “false testimony from wholly fabricated witnesses” and noted that Aarons had made no effort to check the accuracy of the material before filing it.
Court’s Reaction and Sanctions
Beyond the monetary penalty, the supreme court criticised Aarons for showing a lack of remorse and little concern for his client’s defence. The panel ordered that he be referred to the state attorney disciplinary board for further investigation. During a hearing on 21 August, Justice Shannon Bacon pressed Aarons on his awareness of AI’s tendency to hallucinate, asking whether he followed news or radio reports about the technology’s pitfalls. Aarons responded that he regretted the mistake but hoped the board would consider it an honest error, urging other professionals to treat AI as a powerful yet fallible tool.
Background of the Appeal and Defendant
The brief in question was part of the ongoing appeal of Oscar Renee Sandoval, who was convicted last year of murdering the mother of his children and sentenced to life in prison. Sandoval had pleaded not guilty, and Aarons, a private practitioner based in Santa Fe, was handling the appeal before the court reassigned the case to public defender Kim Chavez Cook on 2 September. Both Cook and the local district attorney’s office declined to comment on the matter.
Why it Matters
This case underscores a pressing challenge for the legal profession: as AI tools like ChatGPT become more accessible, lawyers must balance efficiency with rigorous fact‑checking. The New Mexico supreme court’s firm stance sends a clear signal that courts will not tolerate unverified AI‑generated content, reinforcing the duty of counsel to ensure every submission is accurate and trustworthy. For technologists and legal practitioners alike, the episode serves as a cautionary tale that innovation must be paired with vigilance, lest the very tools meant to aid justice inadvertently undermine it.
