Senator Grassley Calls on the Federal Judiciary to Formally Regulate AI Use (www.judiciary.senate.gov)

Sen. Chuck Grassley on the Senate floor urged the federal judiciary to formally regulate and restrict the use of generative AI in judges’ chambers after his oversight uncovered two published court orders riddled with fabricated and misquoted material that staff had generated with AI. The flawed orders—one naming nonparties, misquoting statutes and citing non-existent evidence; the other containing fake case quotes and misattributions—were later removed from the public record. Grassley condemned the errors as unacceptable, emphasized that Article III judges (not software) are responsible for judicial decisions, and warned that if the judiciary does not act, Congress will step in. Technically and institutionally, the episode highlights major risks from unvetted AI: hallucinations, unreliable citations, and the potential to distort rights and precedents—risks disproportionately harmful to unrepresented or indigent litigants. Grassley called on the Administrative Office of the U.S. Courts and the Judicial Conference to issue decisive guidance, and praised localized rules like SDNY Judge Cronan’s requirement that attorneys disclose AI use and personally certify filings. The likely implications include new chamber-level verification protocols, mandatory AI disclosure, training, potential discipline for misuse, and increased congressional oversight if the judiciary fails to implement robust safeguards.