New Mexico Attorney Fined for Using AI-Fabricated Testimony

The New Mexico Supreme Court has fined attorney Stephen Aarons $5,000 for incorporating AI-generated fabrications into a legal brief. According to a report from Reuters, the disciplinary action stems from an appeal for a murder conviction where Aarons included testimony from non-existent witnesses and inaccurate descriptions of a shooter, all sourced from ChatGPT.

Justice Bacon Challenges Legal Oversight

During an August hearing, Justice C. Shannon Bacon questioned how Aarons remained unaware of the systemic risks associated with artificial intelligence. The Scale Report notes that Justice Bacon pointedly asked the attorney if he engaged with mainstream media or general news, given that the issue of AI hallucinations in legal filings has become a recurring subject of public scrutiny. Aarons admitted he believed the software would provide a robust summary of the trial proceedings, only to present false information to the court.

A Growing Problem in Legal Filings

This incident adds to a troubling trend of lawyers using OpenAI technology to draft documents without adequate verification. Previous cases have seen law firms penalized for submitting briefs containing entirely fabricated citations and misleading precedents. While Aarons has expressed remorse, characterizing his actions as an honest error, the court deemed the failure to verify the factual accuracy of his filings as grounds for contempt.

Legal experts emphasize that this case highlights a critical gap in professional ethics training as generative tools permeate law offices. The reliance on automated summaries without human oversight threatens the integrity of the judicial process, as judges are increasingly forced to dedicate time to verifying AI output rather than focusing on the merits of the case.