A PYMNTS Company

Study Finds 60% of Federal Judges Incorporate AI in Court Work

By  |  March 31, 2026
Blockchain, venture capital, framework ventures, new funding

More than half of U.S. federal judges are now incorporating artificial intelligence into their work, signaling a notable shift in how the judiciary is engaging with emerging technology. According to Reuters, a new study released Monday found that 60% of federal judges reported using at least one generative AI tool in their judicial duties.

    Get the Full Story

    Complete the form to unlock this article and enjoy unlimited free access to all PYMNTS content — no additional logins required.

    yesSubscribe to our daily newsletter, PYMNTS Today.

    By completing this form, you agree to receive marketing communications from PYMNTS and to the sharing of your information with our sponsor, if applicable, in accordance with our Privacy Policy and Terms and Conditions.

    The research, conducted by Northwestern University, is believed to be the first random-sample study examining AI adoption among federal judges since the surge in legal tech use following the 2022 debut of ChatGPT from OpenAI. Per Reuters, the study highlights both growing interest in AI and lingering caution among judges.

    While a majority of respondents indicated at least occasional use of AI tools for tasks such as legal research, document review, and drafting, only 22% said they rely on the technology regularly, either daily or weekly. Meanwhile, 38% reported they have never used AI in their judicial roles, underscoring uneven adoption across the bench.

    “Judges see the potential ⁠here,” said Northwestern law professor Daniel Linna, one of the study’s co-authors. “Of course there’s risks, but there are ⁠benefits if we get this right.”

    The findings come amid ongoing concerns about the reliability of AI in legal settings. According to Reuters, several lawyers have faced sanctions for submitting court filings that cited fabricated cases generated by AI systems. In addition, two judges acknowledged last year that staff members used AI tools to help draft court orders that later contained errors.

    The study drew on responses from 112 federal judges, selected from a broader random sample of 502 judges across bankruptcy, magistrate, district, and appellate courts. Researchers say this sample size is sufficient to identify broader trends in judicial attitudes toward AI, per Reuters.

    Judges reported using a range of tools, including general-purpose large language models like ChatGPT and Claude, as well as specialized legal AI platforms such as CoCounsel, Westlaw AI-Assisted, Deep Research, and Lexis+ AI. Many of these tools are developed by companies under the umbrella of Thomson Reuters, Reuters’ parent company.

    The study found that judges tend to favor legal-specific AI tools over general-purpose systems. Legal research emerged as the most common application, with 30% of respondents citing it as their primary use case. Document review followed at nearly 16%.

    Policies governing AI use vary widely across judicial chambers. About one-third of judges said they either permit or actively encourage AI use, while 20% have formal bans in place. Another 18% discourage AI use without outright prohibiting it, and 24% reported having no formal policy at all.

    Training also appears to be a significant gap. More than 45% of judges said they had not received any AI-related training from court administration, a finding that could influence future decisions on funding and education within the judiciary.

    Source: Reuters