US Lawyer Apologizes for Fabricating ChatGPT-Generated Cases

New York-based lawyer Steven Schwartz has issued an apology to a judge after using OpenAI’s ChatGPT, an advanced language model, to prepare a crucial court filing, only to discover that the programme generated fabricated cases and rulings. The incident took place during the proceedings of Roberto Mata‘s civil case at the Manhattan federal court, where Mata is suing Colombian airline Avianca. Schwartz filed a response to Avianca’s lawyers’ request to dismiss the case with claims citing judicial decisions, including Petersen v. Iran Air, Varghese v. China Southern Airlines and Shaboon v. Egyptair. However, neither Avianca’s lawyer nor the presiding judge could locate those cases, leading to Schwartz admitting that ChatGPT had fabricated the entire response. Schwartz expressed his regret and apologized for the incident, which he claimed was not his intention to mislead the court. This incident proves that AI tools still rely on human supervision and judgement.