UC Berkeley law bans AI in coursework

UC Berkeley School of Law has adopted a new default rule that bars students from using artificial intelligence for work submitted for credit, adding one of the clearest restrictions yet by a major U.S. law school. The policy, posted by Berkeley Law and described by Forbes and Business Insider on May 22, takes effect in Summer 2026. It says students may not use AI to conceptualize, outline, draft, revise or edit graded work unless an instructor explicitly allows it. The rule also bars AI use in exams by default. ### What exactly does Berkeley Law’s new rule prohibit? Berkeley Law’s published policy says the default ban covers AI use “in connection with work submitted for credit” and is meant to preserve students’ ability to do core legal reasoning themselves. The school says students need to develop the skills to conceptualize, outline, draft, revise and edit their own work, even as AI becomes more common in legal practice. (law.berkeley.edu) The Summer 2026 policy also says AI may not be used for any purpose in exam situations unless a faculty member specifically authorizes it. Berkeley Law added that some courses may depart from the default rule, particularly classes designed to teach ethical and effective use of AI in legal work. ### Why did the faculty say they moved now? (law.berkeley.edu) Professor Chris Hoofnagle, who helped write the policy, told Business Insider the rule was designed with first-year students in mind and was meant to protect the “value add of a lawyer.” He said the school wanted students to build the fundamental skills required for “AI lawyering,” rather than rely on systems that can generate a paper from start to finish. (law.berkeley.edu) The policy text itself says “thinking remains the sine qua non of good lawyering” and says AI use must be paired with the cognitive skills needed to deploy the technology strategically, assess its output critically and meet ethical obligations to clients and the legal system. ### Is this a total ban on AI at Berkeley Law? The Berkeley rule is not an absolute prohibition on AI across the law school. (aol.com) The posted policy says it is a default position for graded work and exams, while allowing instructors to authorize AI use in particular courses, assignments or exercises. Berkeley Law’s broader academic pages also continue to describe AI-focused coursework and resources for students studying technology and law. (law.berkeley.edu) Forbes reported the school’s move as a stricter turn from earlier guidance, and the American Bar Association Journal said the new policy bans AI for class assignments and during exams unless otherwise permitted. ### How does the school describe the tension between training and technology? (law.berkeley.edu) Berkeley Law’s policy says future lawyers may need to use AI fluently, but argues that current tools still require users to have enough legal judgment to evaluate the output and use it responsibly. That framing places the school’s restriction on graded work alongside continued recognition that AI will be part of legal practice. (forbes.com) Business Insider’s account of Hoofnagle’s comments points in the same direction. He said the issue was not banning the technology outright, but making sure students first learn the underlying reasoning and writing skills the profession expects. ### Where can students and faculty see what happens next? The Berkeley Law policy page says the rule is effective in Summer 2026 and is now part of the school’s academic rules. (law.berkeley.edu) The full text is posted on Berkeley Law’s website, where any course-level exceptions would be reflected through instructor guidance and assignment-specific permissions. (law.berkeley.edu) (aol.com)