The New Mexico Supreme Court has ordered attorney Stephen Arons to pay a 5,000 dollar fine and has found him in contempt of court for submitting documents containing AI-fabricated testimony. The state's highest court also referred Arons to the disciplinary committee, stating that he "failed to verify the factual assertions and legal grounds in an AI-generated document."
New Mexico Supreme Court Fines Lawyer 5,000 Dollars for Citing False ChatGPT Testimony
This article is a translation. Read the Japanese original
According to court records, the appellate brief submitted by Arons included false testimony from non-existent eyewitnesses and police officers, as well as incorrect descriptions regarding the defendant's clothing and appearance. Arons admitted that while using ChatGPT (an AI service provided by OpenAI) to create summaries by inputting trial records, the AI generated erroneous information. He stated that he assumed it would produce accurate summaries, given that AI is widely used in the legal and medical fields.
Regarding this case, Judge C. Shannon Bacon criticized Arons' claim that he was unaware of the risk of AI hallucinations (a phenomenon where AI generates information that is inconsistent with facts), noting that it is an issue frequently reported in news and radio. The judge pointed out that lawyers have an obligation to verify the accuracy of any information they sign, regardless of the tools they use.
Sources
- Lawyer fined $5K over AI-hallucinated witnesses in a murder case （The Verge AI, 2026-09-11）
- ChatGPT-using lawyer punished for citing fake testimony from made-up witnesses （Ars Technica AI, 2026-09-11）