A New York attorney made the unwise choice to trust his legal research to ChatGPT and now faces a court hearing of his own. Steven Schwartz's firm, Levidow, Levidow & Oberman, was representing a client suing Colombian airline Avianca for injuries sustained during an encounter with a serving cart when it submitted documents citing relevant court decisions against other airlines, such as "Miller v. United Airlines," "Shaboon v. EgyptAir," and "Varghese v. China Southern Airlines," per Quartz. The problem was that no one could find these cases. "Six of the submitted cases appear to be bogus judicial decisions with bogus quotes and bogus internal citations," Judge Kevin Castel wrote in an order demanding an explanation, per the BBC.
In a sworn affidavit submitted Thursday, Schwartz---a colleague of the plaintiff's lawyer---said he "greatly regrets" using ChatGPT for research. He added it was his first time and he was therefore "unaware of the possibility that its content could be false," the New York Times reports. Lawyers representing Avianca began to suspect AI involvement after they tried and failed to find the cited cases. Schwartz's firm was apparently so confident in the lawyer's findings that it submitted a list of docket numbers, dates, courts that had heard the cases, and judges who'd allegedly overseen them, per the Times. ChatGPT had invented it all, though in Schwartz's defense, it wouldn't fess up to the deception.
A screenshot submitted to the court shows the chatbot telling Schwartz that "Varghese v. China Southern Airlines" is a real case. The chatbot even claims to have double-checked its sources, finding the case in legal reference databases including LexisNexis and Westlaw, per the BBC. Schwartz, who filed the lawsuit in state court before it was moved to Manhattan's federal court, where he is unable to practice, said the colleague who took over had no idea about his research methods. In his affidavit, the lawyer, Peter LoDuca, said he had "no reason to doubt the sincerity" of Schwartz's work. Both lawyers now face possible sanctions at a June 8 hearing, per the BBC. (Researchers predict AI bots will become "standard tools" for lawyers.)