Attorney Steven Aarons has been fined $5,000 for submitting false testimonies created by ChatGPT in a murder case, raising concerns about AI use in legal proceedings.
On September 9, attorney Steven Aarons faced a $5,000 penalty for presenting fabricated testimonies generated by the ChatGPT chatbot in appellate materials. This decision was made by the New Mexico Supreme Court.
Aarons represented Oscar Rene Sandoval, who received a life sentence in February 2025 for the murder of Shirin Al-Jibouri, alongside an additional four and a half years for armed assault on a witness.
Details of the Fabricated Evidence
Reports indicate that Aarons used ChatGPT to summarize trial transcripts and other case documents, subsequently filing the AI-generated summary in court in August 2025.
Shortly thereafter, the prosecution sought to exclude portions of the document due to inaccuracies.
The submitted materials included testimonies from four fictitious witnesses, including two police officers, and fabricated claims attributed to actual case participants regarding threats made by Sandoval, as well as incorrect descriptions of the shooter's appearance.
Additionally, the document misrepresented the content of two legal precedents.
Aarons acknowledged that he submitted the document without fact-checking or informing his client about the inaccuracies or the ensuing investigation into his actions.
In comments to Reuters, he expressed surprise at AI's capacity to create false information.
Consequences of the Court's Ruling
The court found Aarons in contempt and has referred the case to the disciplinary committee, suspending his ability to practice before the state Supreme Court pending the investigation's outcome.
The fine is due to be paid to the state's client protection fund within 30 days, and all previously submitted documents have been excluded from the case. A new public defender has been appointed to handle the appeal, which is set to proceed in the 2026-2027 court term.
During the proceedings, judges stressed that attorneys bear the responsibility for verifying the accuracy of documents they submit, regardless of the tools employed. Aarons later remarked, “I regret this, but I hope the disciplinary committee will consider that it was a good faith mistake.”
This incident follows a similar case in May 2023, when attorney Steven Schwartz submitted fabricated ChatGPT-generated precedents in a lawsuit against Avianca Airlines, also claiming ignorance regarding the chatbot's potential to produce false information.
