Several questions remain open. The column does not provide full detail on either ruling, and this account does not include the opinions themselves. Practice norms are also unsettled. A Northwestern survey of 502 federal judges found that over 60 percent use at least one AI tool in chambers, mostly for legal research and document review. A Relativity-NAWJ survey found that about one in five judges prohibit court staff from using AI, while 7.4 percent of federal and 14.2 percent of state judges permit and encourage it. State policies vary, and in August 2026 the Supreme Court declined to adopt a categorical ban on judicial AI use.
For attorneys, the practical signal is that courts are setting standards now. Pennsylvania's Judiciary Project for judges has expanded to nearly 25 chambers, including courts in Delaware, New Jersey, Maryland, and the U.S. Court of Appeals for the Third Circuit. California requires judges using generative AI for any adjudicative task to take reasonable steps to verify its accuracy. Lawyers should expect scrutiny of AI-drafted filings, and anyone relying on AI output in an arrest, charging, or evidentiary decision should be prepared to show independent human verification.