SANTA FE — September 13, 2026. The New Mexico Supreme Court fined defense attorney Stephen Aarons $5,000 after an appeal brief he submitted contained fabricated police testimony and witnesses that did not exist.
Aarons told the court that he used ChatGPT to summarize trial materials while preparing an appeal for a client convicted of murder. He said he did not understand how confidently an artificial-intelligence system could invent facts. The justices held him in contempt for failing to verify the filing and referred the matter to a disciplinary board.
The court said the brief included false testimony from entirely fictional witnesses. The defendant’s appeal remains pending and has been reassigned to a public defender.
Why this case stands out
Judges have sanctioned a growing number of lawyers for filing AI-generated briefs containing nonexistent court decisions or inaccurate quotations. This case went further by inserting invented factual material into a criminal appeal, where errors could directly affect someone facing a life sentence.
Generative AI predicts plausible language. It does not automatically distinguish between verified evidence and a convincing invention. A name, quotation or citation produced by an AI system must be checked against the original record before it is used professionally.
Responsibility remains human
Aarons described the incident as an honest mistake and a lesson for professionals. The court nevertheless held the attorney responsible because lawyers have an independent duty to verify filings submitted in a client’s name.
The same lesson applies to journalism: no AI-generated witness, quotation, statistic or citation should be published until a human checks the primary source.
Source: Reuters, September 11, 2026.