New Mexico Lawyer: $5,000 AI Contempt Fine for Fabricated Testimony

• New Mexico attorney Stephen Aarons was held in contempt and fined $5,000 for filing an appellate brief with AI-generated fabricated witness testimony.
• Aarons used ChatGPT to summarize trial materials for Oscar Renee Sandoval's murder appeal, expecting a 'bulletproof summary' but receiving imaginary facts.
• The New Mexico Supreme Court expressed dismay at Aarons' reliance on AI without verification and referred him to the attorney disciplinary board. He was also barred from appearing before the New Mexico Supreme Court while disciplinary proceedings are pending.
• Aarons claimed he misunderstood AI's hallucination capabilities, but the court cited his 'lack of remorse and a lack of concern for his client'.
• Sandoval's appeal is ongoing but has been reassigned to a new public defender, underscoring the necessity for lawyers to verify all AI-generated content.

This case serves as a stark warning that lawyers must rigorously verify all AI-generated content, especially factual claims and citations, before submission to court.

A New Mexico lawyer, Stephen Aarons, has been held in contempt of court and ordered to pay a $5,000 fine after submitting an appellate brief that contained entirely fabricated witness testimony. The New Mexico Supreme Court found that the brief included "false testimony from wholly fabricated witnesses," a significant breach of professional conduct.

Mr. Aarons was representing Oscar Renee Sandoval, who had been convicted of murdering the mother of his children and sentenced to life in prison. In preparing for Sandoval's appeal, Aarons utilized ChatGPT to summarize a computer-generated transcript and other case materials. He informed the court that he had anticipated the chatbot would produce a "bulletproof summary" of the proceedings. However, the AI-generated content introduced imaginary testimony, non-existent witnesses, and even specific, fictional details, such as "fictional statements that the shooter was wearing dark pants and a white shirt," purporting to alter the factual record of a murder appeal involving a client serving a life sentence.

The justices expressed considerable dismay that Aarons had apparently relied on ChatGPT's output without verifying the existence of the individuals or the testimony it described. During an August hearing, Justice C. Shannon Bacon highlighted the widespread awareness of AI's propensity for "hallucinations," questioning Aarons' apparent lack of familiarity with these well-publicized issues. This incident underscores the critical need for human oversight in AI-assisted legal work, especially when dealing with factual claims and citations.

Across the country, lawyers have faced sanctions for submitting AI-generated briefs containing fabricated cases, spurious quotations, and non-existent citations. Despite this trend, Aarons reportedly told Reuters that he had not fully grasped the extent to which AI could invent facts, characterizing the event as an "honest mistake" and a "lesson learned for all professionals who rely upon this powerful but sometimes unstable technology." However, the New Mexico Supreme Court concluded that Aarons had demonstrated "a lack of remorse and a lack of concern for his client." Consequently, in addition to the contempt finding and the $5,000 New Mexico lawyer AI contempt fine, the court referred Aarons to the attorney disciplinary board for further investigation, initiating a lawyer AI disciplinary referral. He was also barred from appearing before the New Mexico Supreme Court while these disciplinary proceedings are pending.

The repercussions of this incident extend directly to Oscar Renee Sandoval's murder appeal. While the appeal remains pending, it has since been reassigned to New Mexico public defender Kim Chavez Cook, highlighting the disruption caused by the original counsel's actions. The case serves as a stark warning that lawyers must rigorously verify all AI-generated content, especially factual claims and citations, before submission to court.

This Stephen Aarons AI sanction reinforces the fundamental principle that artificial intelligence, while a potentially useful tool, cannot substitute for a lawyer's obligation to meticulously review the record and confirm the veracity of every piece of information presented to a court. The responsibility to ensure that every quoted witness is a real person who actually testified, and that every word submitted is accurate, rests squarely with the attorney. Failure to do so can result in severe consequences, including contempt charges, financial penalties, and disciplinary referrals, as demonstrated by this New Mexico Supreme Court AI ruling.

This case serves as a stark warning that lawyers must rigorously verify all AI-generated content, especially factual claims and citations, before submission to court. Failure to do so can result in severe sanctions, including contempt charges, fines, and disciplinary referrals, underscoring the critical need for human oversight in AI-assisted legal work.

Source: Original reporting via Reuters