NEW YORK — In an unusual turn of events during a recent court hearing, the judges of a New York appeals court discovered that the individual presenting an argument via video was not only unqualified but entirely fictional. This incident occurred on March 26 at the New York State Supreme Court Appellate Division’s First Judicial Department, where Jerome Dewald, the plaintiff in an employment dispute, attempted to make his case.
“The appellant has submitted a video for his argument,” announced Justice Sallie Manzanet-Daniels before playing the clip. On screen appeared a polished avatar — a young man with a stylish haircut, dressed in a button-down shirt and sweater.
“May it please the court,” the avatar began, “I come here today a humble pro se before a panel of five distinguished justices.”
However, the mood shifted when Justice Manzanet-Daniels interrupted, asking, “Is that counsel for the case?” Dewald promptly replied, “I generated that. That’s not a real person.”
The judge expressed her displeasure, stating, “It would have been nice to know that when you made your application. You did not tell me that, sir,” before demanding that the video be turned off. “I don’t appreciate being misled,” she added, allowing Dewald to continue with his argument.
Dewald later apologized to the court, clarifying that he had no intention to mislead. Without legal representation, he sought to use the avatar to present his arguments more effectively than he felt he could do himself. He utilized a product from a San Francisco tech company to create the avatar after failing to generate a digital likeness of himself in time for the hearing.
Reflecting on the court’s reaction, Dewald acknowledged, “The court was really upset about it. They chewed me up pretty good.”
The use of AI in legal settings has raised eyebrows before. In June 2023, two attorneys and a law firm faced $5,000 fines from a federal judge in New York after they cited fictitious legal cases generated by an AI tool during their research. The firm admitted to making a “good faith mistake” in assuming that AI could not fabricate information.
Later that year, similar errors occurred when lawyers for Michael Cohen, former personal attorney to President Donald Trump, inadvertently cited fabricated court rulings created by an AI tool during legal proceedings. Cohen accepted responsibility, stating he was unaware that the Google tool he was using was capable of producing such inaccuracies.
In contrast to these mishaps, Arizona’s Supreme Court recently embraced AI by employing two avatars, named “Daniel” and “Victoria,” to summarize court rulings for public understanding.
Daniel Shin, an assistant director of research at William & Mary Law School’s Center for Legal and Court Technology, noted that Dewald’s experience was predictable. “From my perspective, it was inevitable,” he said, emphasizing that while lawyers are unlikely to take such risks due to professional standards and potential disbarment, individuals representing themselves often lack guidance on the implications of using synthetic video presentations.
Dewald mentioned he strives to stay informed about technology trends, having recently attended an American Bar Association webinar discussing AI in the legal field. As of Thursday, his case remains pending before the appeals court.