The legal technology company Clio is seeking to integrate artificial intelligence (AI) into the judiciary, aiming to create tools that streamline judges’ workloads. A comprehensive survey of state court policies indicates a wide variation in whether judges are permitted to use AI, reflecting divergent regulatory stances across the United States.
Central to Clio’s initiative are strategic hires, including Pablo Arredondo, cofounder of Casetext, who has taken on a leadership role as senior vice president, judiciary. Bridget M. McCormack, former Michigan Supreme Court chief justice, now serves as Clio’s general manager for judiciary operations. These appointments underscore Clio’s commitment to revolutionizing judicial operations through AI.
However, the integration of AI in judicial processes faces significant regulatory challenges. For instance, Massachusetts adopts the most stringent approach, outright banning AI use for judicial tasks such as legal research and writing, though permitting it for certain administrative functions. This is formalized through their “Interim Guidelines for use of Generative AI,” which have not been updated since their adoption in November 2025.
In contrast, other states exhibit more lenient policies. New York, for example, allows the use of approved AI tools in specific contexts, supported by judicial ethics opinions that validate AI-generated summary use, provided judges maintain ultimate responsibility over these functions.
Despite the range of state policies, a consistent theme is that AI cannot substitute a judge’s decision-making role. AI tools may assist in administrative functions but must not replace the judicial responsibility to independently verify all research and judgments. Additionally, states impose strict confidentiality measures, prohibiting the use of nonpublic or sensitive information in AI applications.
The potential for AI to assist in routine judicial tasks such as scheduling or note-taking is recognized, yet the core debate centers around more substantive applications, such as analyzing evidence or drafting opinions. As Clio leads the charge into this market, other technology providers may follow, which places a spotlight on the evolution of policies that govern AI’s role in the judiciary.
The 50-state survey reveals no national consensus, ranging from the outright prohibition in Massachusetts to the more advisory and less restrictive guidelines seen in states like Rhode Island and Pennsylvania. This variability highlights the need for an ongoing dialogue within the legal community as it navigates this complex landscape.