AI-Fabricated Testimony Penalty
New Mexico Supreme Court Fines Defense Lawyer Over AI-Fabricated Testimony
Anonymous
Key Points
1
The $5,000 fine was directed to New Mexico’s client protection fund, the court struck the defective briefs, and Sandoval’s appeal was ordered to restart from the beginning.
2
Aarons, who has more than 40 years of experience as a Santa Fe defense attorney, acknowledged at an August 21 hearing that it was “stupidity” to rely on the AI-generated material without verifying it.
3
Aarons said the episode should serve as a broader warning, calling it “a lesson learned for all professionals who rely upon this powerful but sometimes unstable technology.”
4
The case is part of a wider pattern in which dozens of lawyers have been sanctioned for AI-generated legal errors, including fabricated case citations and misquoted law; the New Mexico filing was notable because it allegedly fabricated witness testimony in a criminal appeal.
5
OpenAI did not immediately respond to a request for comment about the court’s findings and the use of ChatGPT in the filing.
A New Mexico defense attorney has been fined $5,000 and held in contempt of court after submitting a murder-appeal brief written with ChatGPT that included fabricated police testimony and nonexistent witnesses The Verge. Stephen Aarons, a Santa Fe lawyer with over 40 years of experience, failed to verify the AI-generated material before filing it on behalf of his client, Oscar Renee Sandoval The Guardian. The court struck the defective briefs and ordered the appeal to restart from scratch with a public defender assigned to the case.
How AI Generated Fake Evidence in Court
The brief Aarons submitted to the New Mexico Supreme Court contained invented details that never happened. ChatGPT fabricated claims about what the shooter wore — dark pants and a white shirt — and created nonexistent witnesses to support the defense Yahoo News. Aarons did not catch these false facts before filing the document with the court, putting his client's case at serious risk.
At an August 21 hearing, Aarons admitted the mistake was "stupidity" on his part Newscord. He said he had not understood how extensively AI tools could invent facts and called the incident an honest mistake. The court, however, was not satisfied with his explanation.
Court's Harsh Judgment on Negligence and Lack of Remorse
The New Mexico Supreme Court criticized Aarons for showing little remorse or concern for his client's predicament Gadget Review. The $5,000 fine was directed to the state's client protection fund as compensation for harm caused. Beyond the fine, the court removed Aarons from the appeal entirely and referred his conduct to the state disciplinary board for further investigation.
This case is part of a broader pattern. Dozens of lawyers across the country have been sanctioned for submitting AI-generated legal work filled with errors Newscord. Most cases involved fabricated case citations or misquoted law, but Aarons' filing was unusual because it invented witness testimony in a criminal murder appeal — a much more serious breach.
A Warning for Professionals Using AI Tools
Aarons said the case should serve as a cautionary tale for all professionals who use AI. He called it "a lesson learned for all professionals who rely upon this powerful but sometimes unstable technology" Newscord. His plea echoes growing concerns among judges and legal experts about AI's ability to confidently generate false information — a phenomenon known as "hallucination."
OpenAI, the company behind ChatGPT, did not respond to requests for comment about the court's findings or how the tool was used in Aarons' filing. The case underscores the need for lawyers to independently verify any AI-generated legal material before submitting it to court.
Sentiment
Mostly NegativeNEGATIVENEUTRALPOSITIVE
Publishers
16
Articles
6
Reach
22
A
AnonymousAuthor
