What happened
PYMNTS reports that several U.S. law schools are restricting generative AI and, in some classrooms, student devices. The University of Chicago Law School reportedly barred phones, tablets and laptops from core first-year classrooms, while UC Berkeley Law reportedly prohibited generative AI assistance on credit-bearing work.
PYMNTS reported on Sept. 7, citing a Financial Times column by Patti Waldmeir, that U.S. law schools have begun restricting artificial intelligence in the classroom. The report identified the University of Chicago Law School and the University of California, Berkeley, School of Law as examples. PYMNTS is the source for this account; the policies were not independently confirmed here.
At Chicago, PYMNTS said phones, tablets and laptops would be forbidden in core first-year classrooms for the new academic year. William Hubbard, chair of the school’s AI committee, reportedly said the purpose was to preserve the difficult, discussion-based learning associated with the Socratic method. The report also said Chicago was not imposing a blanket ban and was considering how to adapt its curriculum to students’ expected AI use.
At Berkeley, PYMNTS reported that a new policy forbids generative AI for conceptualizing, outlining, drafting, revising, translating or editing work submitted for credit. Dean Erwin Chemerinsky reportedly said professors retain some discretion. The source does not provide enforcement procedures, penalties, dates beyond the academic-year reference, or evidence about student compliance.
The restrictions apply to students in the cited classroom and coursework settings, not necessarily to all AI use by those students. No access conditions or pricing apply, and the report does not document any AI product being offered or withdrawn.
Why it matters
The policies reflect a practical tension in professional education: students must develop independent legal reasoning while preparing for a profession that increasingly uses AI. The reported changes also show that institutions are experimenting with narrower classroom and assessment rules rather than adopting one uniform approach. The source does not independently confirm the policies or establish how widely they are being enforced.
Law is an unusually consequential setting for this debate because professional work increasingly incorporates AI while legal education still depends on students learning to read authorities, construct arguments and defend judgments themselves. The reported policies treat those skills as learning objectives that can be undermined when AI performs the core intellectual work.
The Chicago and Berkeley examples suggest two different governance models: controlling the tools present during foundational classroom discussion, and defining prohibited assistance in submitted work. Neither approach, as described by PYMNTS, demonstrates that bans improve learning outcomes. The source supplies no independent study, measurement or comparison of results.
The broader market context in PYMNTS’s article indicates that AI adoption in legal and other professional workplaces is increasing, but its cited investment and enterprise-adoption figures do not establish the effectiveness of classroom restrictions. They mainly explain why schools face pressure to prepare students for AI while limiting dependence on it during training.
What to watch next
Watch whether other law schools adopt device restrictions, define permitted AI assistance more precisely, or redesign assessments around oral examination and supervised work. The important unknowns are how professors interpret exceptions, how compliance is monitored, and whether these restrictions affect students’ preparation for AI-enabled legal practice.
The next meaningful development would be additional named law schools adopting comparable rules or publishing detailed guidance on acceptable uses, disclosure and assessment design.
It is not clear from the report whether Chicago’s device restriction covers every class session, whether accommodations or exceptions exist, or how the school distinguishes AI literacy from prohibited assistance.
It is also unknown how Berkeley’s policy is enforced, whether faculty apply it consistently, and whether students may use AI for ungraded study, research or accessibility purposes.
Any evaluation of these policies should examine learning outcomes and students’ ability to perform legal analysis independently, rather than relying only on adoption claims or anecdotal reactions.