Recently, the Supreme Court of New Mexico handed down a significant penalty to defense attorney Stephen Arens. Arens incorporated unverified information generated by ChatGPT into a criminal appeal document for a murder defendant. This document included fabricated witness statements and false police testimonies, amounting to contempt of court. Consequently, the court imposed a fine of $5,000 on him and referred the matter to the lawyer disciplinary committee for further inquiry. This case represents the first major penalty within the U.S. judicial system for the improper use of AI technology.

Arens contended that he was unaware that AI could generate false information. However, the presiding judge rejected his claim of ignorance regarding the potential for AI to produce inaccurate or hallucinatory content. Over the past year, there have been numerous similar incidents across the United States.

Legal experts assert that this case indicates a 'zero-tolerance' stance by the judicial system towards the misuse of AI technology. The New Mexico Bar Association is currently developing guidelines for the use of AI, mandating that lawyers assume ultimate responsibility for verifying the accuracy of AI-generated content.