A defendant tried to use an AI avatar in a legal appeal. It didn't work

A defendant in a New York appeals court has been slammed by a judge for using an artificial intelligence avatar to represent himself in a recent case. 

A New York appeals court faced an unusual situation in late March when Jerome Dewald, representing himself in an employment dispute, submitted an AI-generated avatar to present his legal arguments via video, a livestream of the hearing shows. 

It’s the latest example of artificial intelligence tools trickling their way into courtrooms. 

Within seconds of the video starting, Justice Sallie Manzanet-Daniels called for it to stop, asking whether the avatar was counsel for the case. 

“I generated that,” 74-year-old Dewald responded, adding, “That is not a real person.”

The judge appeared displeased, retorting, “It would have been nice to know that when you made your application,” stating that the defendant had previously appeared before the court and been able to testify verbally in the past. 

“I don’t appreciate being misled,” the judge added. 

She asked the defendant if he was suffering from an ailment that prevented him from articulating before adding, “You are not going to use this courtroom as a launch for your business,” and then yelling, “Shut that off,” pointing to the video screen. 

Appellate Division, First Department Stream. Source: YouTube

Dewald later apologized, explaining he thought the AI avatar would deliver his arguments more eloquently than he could.

Speaking to The Associated Press, Dewald said he applied to the court for permission to play a prerecorded video, then used a San Francisco tech company to create the AI avatar. 

He originally tried to generate a digital replica of himself but was prevented by time constraints before the hearing. “The court was really upset about it,” Dewald conceded, adding, “They chewed me up pretty good.”

Related: Meta's Llama 4 puts US back in lead to ‘win the AI race’ — David Sacks

The incident highlights growing challenges as AI enters the legal world. 

In 2023, a New York lawyer was blasted for citing fake cases generated by ChatGPT in a legal brief as part of a lawsuit against a Columbian airline.

In March, Arizona’s Supreme Court began using two AI-generated avatars, similar to the one that Dewald used in New York, to summarize court rulings for the public.

In September, the US Federal Trade Commission took action against companies it claimed misled consumers using AI, including a firm that offered an AI lawyer.

Magazine: ‘Chernobyl’ needed to wake people to AI risks, Studio Ghibli memes: AI Eye