UC Berkeley Law School Tightens Student AI Use Policy

UC Berkeley Law School approved a stricter AI use policy that, according to Business Insider, forbids students from using AI for conceptualizing, outlining, drafting, revising, editing, translating, or for any purpose in an exam situation. Business Insider reports the policy was approved by a faculty vote, goes into effect this summer, and allows instructors to deviate from the baseline; Business Insider also reports that AI-focused courses will follow different standards.

Business Insider quotes UC Berkeley Law School professor Chris Hoofnagle, who helped draft the policy, saying the policy is "about developing students with the fundamental skills required for AI lawyering." Hoofnagle is quoted in Business Insider as saying the 2023 policy was "too liberal" given the increasing capabilities of generative models, and that such models "can, in effect, write a research paper soup to nuts."

Law-school assessments emphasize demonstration of legal reasoning, statutory analysis, and writing craft. Industry observers note that as LLMs improve in synthesis and drafting, academic programs face a choice between stricter use controls and redesigned assessment formats. For practitioners and instructors, the technical issue is not model capability alone but how reliance on generated text interacts with skills that are hard to measure through take-home assignments.

For legal education specifically, preserving the "value add of a lawyer," as quoted in Business Insider, maps onto curricular debates about what constitutes assessable competence versus automatable work.

• •Whether other top law schools adopt similar faculty-level baseline policies or prefer course-level rules.
• •Enforcement mechanisms and academic-integrity processes for verifying permitted versus prohibited uses.
• •How law curricula and assessment design evolve to test skills that LLMs can assist with versus those that require student-authored reasoning.