Supreme Court justices are also talking about Claude.
Associate Justice Samuel Alito, one of the court’s conservative members, humorously addressed the human-owned large-scale language model during oral arguments before the court on Monday.
“So, I’m just curious, do you think I should ask Claude to judge this case?” Alito asked with a laugh.
The petitioner’s attorney, Adam Unikowski, responded instead that he would “defer to the prudent judgment” of the court.
Screenshot of the oral argument transcript in Juul v. Andre Balazs Properties. Supreme Court of the United States
Claude had little to do with the underlying case, Juul v. Andre Balazs Properties, which concerns the ability of federal courts to confirm arbitral awards.
So why did Alito bring it up?
Unikowski, a lawyer at Jenner & Block who has argued more than a dozen cases before the Supreme Court, also happens to be an AI enthusiast and has been experimenting with the use of technology in his legal practice.
He has posted several newsletters about experiments with AI in law, claiming as far back as June 2024 that AI “can accurately decide cases and write judicial opinions.”
In one instance, Unikowski gave Claude briefs on several Supreme Court cases and found that the large-scale language model “consistently decides the cases correctly,” and that its “dispositions are always reasonable,” even when they differ from how the Supreme Court actually decided the cases.
Unikowski also conducted an experiment last year that allowed Claude to give oral arguments in court, concluding that courts “should allow oral arguments by robot lawyers and should not discourage this practice.”