Judges in a New York appeals court were shocked when they discovered that a man appearing on their video screen wasn’t real. He was an AI-generated avatar created by a man representing himself in a legal case.
Dewald was involved in an employment dispute and had asked to show a video as part of his argument. When the judges watched the video, they saw a polite, well-dressed man begin to speak. But one judge quickly stopped the video and asked who the speaker was. Dewald then admitted that the man wasn’t real — it was an AI avatar.
The court was not pleased. Justice Sallie Manzanet-Daniels said Dewald should have told them about the avatar ahead of time. “I don’t appreciate being misled,” she said, before allowing Dewald to continue his case without the video.
Dewald later apologised. He explained that he didn’t have a lawyer and had trouble speaking clearly, so he used the avatar to present his points more smoothly. He used software from a company in San Francisco to create the digital speaker.
However, not all uses of AI in law have been negative. Arizona’s Supreme Court now uses AI avatars named “Daniel” and “Victoria” to explain rulings to the public on their website.
Experts say using avatars like Dewald’s in serious court matters is risky. Courts have rules and traditions that must be followed, and using fake people without warning can cause confusion and anger.
