AI Governance
Data Centers
Drones
Climate
Privacy
2026 Maine Senate
Bob Dylan
Universities Tighten Student AI Rules
Coverage from Business Insider, Chronicle, and others
Universities and legal-education programs are tightening rules on generative AI in exams, assessed writing, and other credited work.
UC Berkeley Law’s Summer 2026 policy establishes a restrictive default that permits limited source identification but bars AI-assisted drafting and related tasks, while the University of Edinburgh’s philosophy department adopted a default ban on generative AI in assessed work. The broader shift reflects concerns about academic integrity, student development, fabricated citations, confidentiality, and the difficulty of enforcing rules as AI features become embedded in search and legal research tools.
The story has shifted from a general tightening of student AI rules to a more concrete picture of implemented institutional policies, especially Berkeley Law’s restrictive Summer 2026 default and Edinburgh’s assessed-work ban. It also now extends more explicitly into legal practice guidance, with new references to externship-site policies and regulatory pressure on accurate AI use.
The story broadened from two school policy examples into a wider pattern of law-school AI restrictions that now explicitly includes externships and professional confidentiality concerns. The new material also shows implementation pressure from legal-practice guidance and stricter enforcement mechanisms, not just classroom integrity.
The story broadens from Berkeley Law alone to a wider higher-education pattern, as Edinburgh Philosophy adopts a similar default no-AI rule for assessed work. The new emphasis is that stricter AI restrictions are becoming a governance trend across universities, not just a single law-school policy.
- University of Edinburgh Philosophy adopted a default no-generative-AI rule for assessed work.
- Allowed use is narrowed to specified proofreading and citation tools meeting university standards.
- The policy discussion now extends beyond legal education to philosophy assessment.
- Schools can police submissions and exams, but not AI outputs inside research platforms.
UC Berkeley Law has adopted a restrictive generative AI policy for student work, especially exams and credit-bearing assignments. The school says the rule is meant to preserve core legal-writing and analytical skills while still allowing limited, instructor-approved AI use in some contexts. The policy also reflects growing concern about hallucinated citations, enforcement challenges, and the need to teach students how to use AI critically rather than depend on it.