A group of leading US law schools is rewriting how students may use AI, with policies ranging from device-free first-year classes to formal training in legal technology. The important shift is that schools are moving beyond simple bans and deciding when future lawyers should learn to use these systems responsibly.
Watch Desk analysis
What happened
Lawyer Monthly, citing Reuters, reports that at least a dozen US law schools introduced or revised AI policies over the summer. The changes cover classroom devices, examinations, submitted work and compulsory technology training.
The University of Chicago has taken one of the stricter routes for its required first-year curriculum. During the 2026-27 academic year, students will generally be barred from using laptops, tablets and phones in those classes, with limited exceptions such as disability accommodations or instructor-approved activities. Exams will also be held without internet access, electronic files or apps.
That is not a complete rejection of AI. Chicago plans to begin its legal research and writing programme with independent writing before introducing AI for research, revision, draft iteration and oral-argument preparation.
Berkeley Law’s policy, effective from summer 2026, prohibits AI for conceptualising, outlining, drafting, revising, editing or translating assessed work. Students may not upload course materials to generative AI systems, although faculty can set different rules where appropriate.
Columbia Law School is more permissive. Students may use AI for research, criticism, alternative formulations and clarity, but not to perform the original legal reasoning an assignment is designed to assess. AI use in exams is prohibited by default unless an instructor permits it. Columbia is also adding AI training at orientation and in selected first-year courses.
Why it matters
Law schools are trying to solve two problems at once. Students need to develop judgement without outsourcing the thinking that their courses are meant to teach, but graduates will enter a profession where AI tools are increasingly difficult to avoid.
The policy differences also give students and employers a useful map of the emerging argument. Is AI a prohibited shortcut, a supervised research assistant, or a professional skill that should be taught from the start? At present, the answer depends rather heavily on which campus has printed your timetable.
Lawyer Monthly says an archive maintained by Suffolk University Law School dean Andrew Perlman tracks public AI policies at 180 US law schools, with 38 currently listed under mandatory-curriculum requirements. That suggests this is becoming an institutional pattern, not a handful of unusually anxious lecturers guarding the typewriter.
Our read
The strongest policies are likely to be the ones that distinguish between tasks rather than treating “AI use” as one indivisible activity. Preventing a student from outsourcing legal analysis is not the same as forbidding a tool from helping with research or revision.
The practical test will be whether schools can state their rules clearly enough for students to follow them, and whether they teach the limits of AI as seriously as they teach its uses. A ban may protect an assessment. It does not, by itself, prepare anyone for a profession where confidentiality, accuracy, supervision and candour still matter after graduation.
What to watch
- Whether more law schools introduce compulsory AI or technology-competence requirements.
- Whether students receive consistent rules across modules rather than a patchwork of lecturer-by-lecturer exceptions.
- How schools assess unauthorised AI use without creating a new confidence crisis around automated detection.
- Whether legal employers begin treating documented AI competence as a hiring requirement. The central question is no longer whether future lawyers will encounter AI. It is whether education should teach them to use it before the workplace does the teaching badly.
Discussion spark: Should law schools ban AI from assessed work to protect independent legal reasoning, or make supervised AI use part of the qualification because graduates will need that skill immediately?
Sources and evidence
- Law Schools Set New AI Rules for Students – Lawyer Monthly (21 September 2026, 18:46 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.