U.S. Federal Judges Use AI, But Not for Drafting Judgments
Survey Published in 'Sedona Conference Journal'
Six out of Ten Judges Say "I Have Used at Least One AI Tool"
Main Purposes: Legal Research and Document Review
Only 1.8% Use AI for Drafting Judgments or Other Official Documents
A recent survey has found that more than half of U.S. federal judges are utilizing artificial intelligence (AI) in their judicial work. However, AI is being used primarily for supporting tasks such as legal research, rather than for core functions like drafting judgments or making judicial decisions, with only a very small number of judges adopting AI in these critical areas.
According to the report "Artificial Intelligence in Federal Courts: A Random Sample Survey of Judges," published in the U.S. legal journal Sedona Conference Journal, 61.6% of surveyed U.S. federal judges responded that they had used at least one AI tool in their work. This means that approximately six out of ten judges have encountered AI in their work.
Primarily Used for 'Legal Research'
The survey revealed that the areas in which judges and court staff most actively employ AI are 'Legal Research' and 'Document Review.' The highest proportion of respondents indicated use for 'legal research,' with 30.0% of judges and 39.8% of staff identifying this as their primary application. For 'document review,' 15.5% of judges and 16.7% of staff reported using AI. Document review refers to assigning AI to search the contents of specific documents or to answer related questions.
In contrast, only 1.8% of judges reported using AI for core judicial responsibilities such as 'drafting official documents filed in cases (including judgments and orders)' or 'making actual decisions,' each representing a very small minority. Instead, judges showed greater flexibility in using AI for drafting (7.3%) or editing (4.5%) unofficial documents, such as letters or emails. One federal judge emphasized in the survey, "I allow AI as a research assistant in my courtroom, but I have a strict rule that all information provided by AI must be independently cross-checked."
Everyday Use Still Rare... Preference for 'Legal-Specific Tools'
Although the experience of using AI is widespread, few judges use it on a daily basis. In terms of usage frequency, only 5.4% of judges reported using AI daily, and 17.0% weekly, totaling just 22.4% who use it regularly.
Judges also showed a strong preference for legal-specific AI (AI for Law) over general-purpose AI such as ChatGPT. The most widely used individual tool was Thomson Reuters' 'Westlaw AI-Assisted or Deep Research.' OpenAI's ChatGPT (28.6%) and Thomson Reuters' legal AI assistant CoCounsel (20.5%) followed, with Microsoft Copilot (15.2%) and Gemini (15.2%) also making the list.
This survey was conducted by researchers from Northwestern University School of Law, a judge from the U.S. District Court for the Western District of Texas, and the AI & Digital Technology Task Force under the New York City Bar Association. As of August 2025, a random sample of 502 out of 1,738 incumbent U.S. federal judges was surveyed, and 112 responses were collected.
Korean Courts Also Moving Toward 'Coexistence with AI'
The Korean judiciary is also accelerating the adoption of AI. The Sentencing Commission of the Supreme Court plans to conduct a pilot operation in 2028 and officially implement an AI sentencing system from 2030. The Court Administration Office has also piloted a generative AI platform and trial support AI system, built on the judiciary's own infrastructure, in February 2026. This system is also focused on expediting and streamlining legal information searches and reference checks, rather than drafting judgments.
In practice, the number of judges utilizing AI is gradually increasing in Korea as well. However, most use cases are still limited to legal research. One district court judge said, "I occasionally use Gemini for case law research, but due to the court's network separation policy, it is difficult to transmit materials, so I do not use it for other purposes."
Another high court judge stated, "I have used AI such as Claude to research overseas legal trends or foreign judgments for research purposes. While there are not many occasions to use it in general civil or criminal trials, it is definitely useful when handling international matters."
A different high court judge commented, "Services like LBox require ID verification to use AI features. Entering case-related materials into a private company platform and seeking a judgment is not only a security risk but could also violate judicial ethics, so I am reluctant to use it." The judge added, "Although the court has supported subscriptions to LBox and SuperLawyer for a while, most judges only used them for case law searches or keeping up with trends. No judge would ask AI to draft a judgment, unlike lawyers who might."
Hot Picks Today
"Two Spend 500,000 Won in a Flash, Yet Wait Two Hours?" ... 'Dreamland' Shunned by Kids and Foreign Tourists
- The Era of Getting 200 Million Won After 30 Years Is Over... A Japanese Bank's "Perpetual Department Head" Who Just Clocked In Cries Out
- Spain Hit by Its Worst Tragedy... Lee Kangin, Moreno and Real Madrid Join Soccer World in Mourning
- They Said, “Make It Your Home for Life” - Then Suddenly Kicked 200 Passengers Off... What Happened to the Around-the-World Cruise?
- "130 Rice Bowls Sold Out in a Day"... 'Unfiltered Face' Adds 8 Million Won to Night Market Sales: China's Bungeunwoo
Jisoo Kim, Legal Times Reporter
※This article is based on content supplied by Law Times.
© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.