AI-Generated Avatar Sparks Courtroom Clash, Judge Slams Legal Tech Tactic
A 74-year-old tech entrepreneur found himself in hot water inside a New York courtroom after deploying an AI-generated avatar to argue his legal case — without telling the judges. What began as an employment dispute hearing quickly turned into a cautionary tale about the limits of AI in the courtroom.
Courtroom Confusion Over Mysterious 'Lawyer'
The incident occurred on March 26th, when Jerome Dewald — founder of a startup aiming to revolutionise legal self-representation with AI — submitted a video in lieu of an in-person argument. The court had granted him permission to present his case via video due to health-related speaking difficulties. But when the footage played, a strikingly unfamiliar man appeared onscreen.
“Hold on,” said Justice Sallie Manzanet-Daniels, pausing the presentation. “Is that counsel for the case?”
Dewald’s reply drew audible gasps: “I generated that. It’s not a real person.”
The AI-generated speaker, dubbed “Jim” by Dewald, was created using a tool from the avatar platform Tavus. Describing his creation to The Register as a “big, beautiful hunk of a guy,” Dewald explained that the avatar helped him manage the strain of prolonged public speaking. But the court had no prior knowledge that the man on screen was a digital invention.
Judge Pulls No Punches
Justice Manzanet-Daniels did not hide her frustration with the surprise twist. “It would have been nice to know that when you made your application. You did not tell me that, sir, I don’t appreciate being misled,” she said.
Then came the stinging rebuke that has since gone viral: “You are not going to use this courtroom as a launch for your business.”
The judge’s stern words suggest that while technology may be inching into the legal world, there's little patience for AI-driven theatrics that catch the court off guard.
A Growing Tension Between AI and the Legal System
Dewald’s stunt is not the first time AI has stumbled in legal settings. In 2023, a pair of attorneys and their firm were sanctioned for submitting fake case law conjured up by ChatGPT. Earlier this year, the Federal Trade Commission fined the so-called “robot lawyer” company DoNotPay $193,000 for falsely marketing its AI services as equivalent to a human attorney.
These cases highlight a growing discomfort within the justice system around the unchecked use of generative AI. While innovators like Dewald see potential to democratise legal access, the courts remain wary of being used as a stage for unproven tech — especially without transparency.
As the debate over AI’s role in the courtroom rages on, one thing is clear: Justice may be blind, but it isn’t buying the avatar pitch just yet.
technology