UC Berkeley Law’s AI Ban Is a Wake-Up Call for Legal Education
One of the most prestigious law schools in the world has had enough. UC Berkeley Law is rolling out a sweeping AI ban in legal education starting summer 2026, prohibiting students from using artificial intelligence in nearly all graded work. That means no AI for brainstorming, drafting, outlining, writing, revising, translating, or proofreading any submitted assignment. Exams are completely off the table too.
The only carve-out in this AI ban in legal education is using the tools for research purposes, such as locating statutes or case law. But even then, students are fully responsible for every fact they cite. And if fabricated citations show up in their work, that alone is treated as evidence of banned AI use, a policy that directly addresses one of the most well-documented risks of AI in legal settings.
The school put it plainly: “Thinking remains the sine qua non of good lawyering.” The argument is that lawyers who lean on AI before they have built core analytical skills will not be able to use it meaningfully when it counts. They need to know how to think first.
This move comes as AI tools continue to shake up the legal profession on multiple fronts. The technology has shown it can speed up research and drafting considerably, but it has also slipped in errors that go undetected and made mistakes in one out of six cases according to a Stanford study. Bias and the shaping of legal opinions are concerns too.
Berkeley’s policy does leave some flexibility for professors, who can adjust the rules for specific courses, particularly those designed to teach students how to work with AI tools responsibly. But the default position is clear: build the lawyer first, then introduce the machine.
The full policy is available directly from UC Berkeley School of Law.
