Oyez Oyez AI!
AI Order in the Courts - Judicial Use of AI
Adoption of AI in Courts - USA, Intl, States updates
As we navigate the growing AI epoch in mid-stage 2025, artificial intelligence (AI) reshapes our judiciaries.
Courts around the world are beginning to embrace AI to tackle three primary system issues:
staffing shortages,
case backlogs, and
access to justice
»»» Unreported (and not subject to disclosure) is judges’ use of AI tools. In an environment where increasingly, AI is seen as a “Scarlet Letter” for attorneys to use — many being caught out for synthetic citations or cases, somehow the use we know judges (and their clerks) are making of it rides under the radar.
A 2025 survey by the Thomson Reuters Institute and the National Center for State Courts (NCSC) revealed that (only) 17% of U.S. state courts are currently using generative AI (gen AI), with another 17% planning to implement it within the next year.
It should be noted that this is likely a 3x multiple in growth over the previous year.
Overseas it is different. A UNESCO global survey found that 44% of judicial operators across 96 countries are already using AI tools like ChatGPT for work-related tasks, indicating faster uptake internationally.
Morocco openly uses of AI for drafting preliminary judgments in labor and traffic cases, as well as predictive analysis based on historical data.
The U.S. court system is instead exploring AI for targeted use cases, such as chatbots to handle public inquiries outside business hours, aiming to enhance access to justice and achieve cost efficiencies.
Paul Drutz-Hannahs, AI pilot lead at the Administrative Office of the U.S. Courts, emphasized the judiciary's lower risk tolerance compared to private industry, prioritizing public trust.
US State court actions in the U.S. during the first half of 2025 show a patchwork of approaches.
Arizona amended rules to allow "non-production use" of AI for testing purposes.
Arkansas proposed an order prohibiting exposure of internal court data to AI and requiring awareness when using such systems.
Michigan and West Virginia have mandated "tech competency" for judges.
Nevada created an AI guide for judicial officers, explaining generative AI's workings and impacts.
South Carolina issued an interim policy limiting AI use to approved tools under supervision.
Texas concluded that existing evidence rules suffice for handling AI-manipulated evidence.
Vermont released a committee report on AI in disciplinary rules, operations, and court rules.
Virginia amended rules to mandate human oversight in criminal processes using AI.
Washington adopted principles requiring independent judgment and data security when using AI.
These initiatives reflect a growing recognition of AI's role, supported by resources from the NCSC, such as AI policy guides, implementers' forums, and an AI sandbox for experimentation.
61% of U.S. courts anticipate ongoing shortfalls in staffing due to budgetary pullbacks. AI steps right in the gap to automate administrative tasks like transcription, language translation, document processing, and scheduling.
As an example, the Superior Court of Los Angeles County now uses an AI-powered system for jury duty rescheduling via conversational interfaces.
[N.B. We did a 2023 build (and Jan 2024 demo) of a specialized chatbot interface for one 75,000-lawyers strong state bar association for a similar regular time/cost-intensive nagging scheduling item.]
Some estimate AI could save nearly three hours per week now, rising to nine hours in five years, helping address backlogs without replacing human roles.
On a broader scale, AI enhances access to justice through tools like legal advice chatbots, online dispute resolution, and case triaging, particularly benefiting self-represented litigants.
The UN Special Rapporteur on judicial independence notes AI's potential to improve justice administration, including live transcription and translation, while expanding access for underserved groups.
Retired justice John Browning highlights the need for judges to spot red flags like mismatched case numbers, especially in overburdened lower courts.
70% prohibit its use for court business, citing budget constraints and fears of malicious applications like counterfeit evidence.
Compared to the private sector, courts are lagging — yet still moving more rapidly than their customary pace to adopt AI.
+++ ARS TECHNICA AI “Scare-Porn” LINK
