More than 60% of federal judges use AI at work. Some use it to draft the actual ruling.

When Xavier Rodriguez, a Texas-based federal judge with more than 20 years on the bench, prepares for a hearing, he starts by feeding the relevant court filings into an AI tool. The AI produces a timeline of the case and the claims the parties are making. "My law clerks would be wasting 30, 45 minutes, an hour, developing a chronology of events," Rodriguez told The Washington Post. "This thing does it instantaneously." Before hearings, he also asks AI to suggest questions to put to attorneys or identify weaknesses in a plaintiff's argument. In areas where he feels especially confident, he sometimes uses AI to draft the ruling itself.

Rodriguez is not alone. A Northwestern University study published this week — co-authored by Rodriguez — collected responses from 112 federal judges and found more than 60% reported using AI tools at least once in their judicial work. Around 22% said they use AI daily or weekly. The Los Angeles County Superior Court announced a pilot program with Learned Hand, an AI startup for judicial work, and the company said its tool is also being used in trial courts in 10 states and the Michigan Supreme Court.

Rodriguez frames it as a time-saver, not a replacement: "I'm doing my own preparation. I'm not strictly relying on an AI tool. It's just an extra set of eyes." Legal scholars are less relaxed. "Judges, they're responsible for making decisions that are very important to people," said Eric Posner, a law professor at the University of Chicago. "They just can't gamble with a technology that is not fully understood and that is known to hallucinate."

The judiciary has already had one visible AI problem: a separate Washington Post report found federal judges had filed court orders containing AI-generated false quotes and fabricated citations. That was the lawyers. Now the judges are using it too.