Federal judges increasingly turn to AI for case prep and ruling drafts
More than 60% of federal judges use AI tools in their judicial work, according to a Northwestern University study released this week. About 22% use the tools daily or weekly.
Judges are deploying AI across their workflow: analyzing case filings, preparing for hearings, drafting rulings, and identifying legal arguments. Xavier Rodriguez, a Texas federal judge with over 20 years on the bench, said AI generates case timelines in seconds-work that would take his law clerks 30 to 45 minutes or longer.
After deciding how to rule, Rodriguez sometimes uses AI to draft the opinion itself. "I'm doing my own preparation," he said. "It's just an extra set of eyes."
Courts are formalizing the shift. The Los Angeles County Superior Court launched a pilot program in March with Learned Hand, a legal startup building AI tools for judges. Thomson Reuters and LexisNexis have contracts to provide AI tools to the federal judiciary. Learned Hand's system is already in use across trial courts in 10 states and the Michigan Supreme Court.
Speed versus reliability
Judges say the time savings matter for an overburdened system. Christopher Patterson, chief judge of Florida's 14th Circuit Court, noted that court filings continue to rise. "We've got to find a way to meet that need in an ethical, responsible way," he said.
Samantha Jessner, a Los Angeles County Superior Court judge testing Learned Hand's tool, uses it to identify controlling authorities and analyze whether claims have legal backing. "You can interface with the tool and ask the tool pretty specific questions," she said.
But reliability remains a concern. A 2024 Stanford University study found that Thomson Reuters and LexisNexis legal AI tools made mistakes in 17% to 33% of queries-better than general-purpose chatbots, but still problematic for judicial decisions.
Last year, hallucinations appeared in court filings by two federal judges. Henry T. Wingate of the Southern District of Mississippi and Julien Xavier Neals of the District of New Jersey cited nonexistent cases and attributed false quotes to plaintiffs. Both judges blamed clerks and interns using AI. Neals subsequently banned generative AI from his chambers.
Judges maintain final authority
All judges interviewed said they do not let AI decide how they will rule. The tools serve as drafting aids and research assistants, not decision-makers.
Patterson emphasized the point to his staff: "Don't let (AI) substitute your judgment for the tool's judgment. You can't let that happen."
Eric Posner, a law professor at the University of Chicago, warned that judges cannot afford to gamble on unreliable technology. "Judges are responsible for making decisions that are very important to people," he said. "They just can't gamble with a technology that is not fully understood and that is known to hallucinate."
Michael Navin, a consultant at the National Center for State Courts, said adoption is real but still early. "I don't get a sense that there's an overwhelming use of AI by judges, but it is there."
Courts are proceeding cautiously. The Los Angeles pilot will run for a year and is limited to certain civil motions. Florida's court is testing AI on both old and active cases before deciding on wider use.
Legal AI vendors say their tools are less prone to hallucination because they source answers from databases of court cases and legal documents rather than the open internet. LexisNexis and Thomson Reuters said in statements that their tools have improved since the Stanford study and link back to source cases so users can verify information.
For more on AI for Legal professionals, or explore how AI Learning Path for Paralegals covers document review and legal research automation.
Your membership also unlocks:
