The University of Chicago Law School is going laptop-free for first-year students this fall. In a move designed to curb AI-enabled cheating, the school requires laptops closed during in-class sessions and stacks additional guardrails around major assessments.

Dean Adam Chilton, the law school’s dean, says the goal is not to pretend AI can be turned off. It’s to build “space for both of these modes of learning,” where students learn to think for themselves without AI first, and then learn to use AI ethically once they have the foundation.

That tension is no longer theoretical. Business Insider previously reported that Brown University disciplined dozens of students after uncovering what administrators described as a widespread AI-assisted cheating scandal. The pattern matters: when take-home assignments and remote assessments become easier to outsource to a chatbot, educators lose something far more important than a grade. They lose evidence that a student can actually reason, argue, and explain.

UChicago’s response is basically a curriculum and assessment redesign, not a technological whack-a-mole. The school’s new strategy includes laptop-free first-year classes, in-person proctored exams that prevent students from accessing outside materials, and oral defenses for major research papers so students can explain and defend their work. In plain terms, the school is shifting from “show us the output” to “show us the thinking process.” If the output was generated with AI, the defense is where students get measured on whether they can stand behind it.

But the school also acknowledges the other reality driving this change: AI is becoming part of the legal profession. Chilton points out that law firms increasingly expect new hires to use technology efficiently and responsibly. That expectation makes a pure ban on AI unrealistic. Even if a school tries to lock down the classroom, students are still entering a profession where AI tools are increasingly baked into day-to-day workflows.

So the strategy tries to separate skills. Chicago Law says it is redesigning its curriculum to distinguish between what students should develop independently and what can be embraced with AI. Chilton criticizes take-home assignments that allow students to complete work with AI “without thinking for yourself, without learning for yourself.” His point is not subtle: if the assessment method rewards outsourcing, the curriculum becomes less about mastery and more about access.

The school’s plan goes beyond proctoring and oral defenses. It also expands AI instruction. UChicago is integrating AI into legal writing courses, adding more AI-focused classes, and providing students access to legal AI tools, including Harvey and Legora. That matters because it reflects a broader institutional problem: educators cannot address AI’s impact using only restrictions. Students also need structured, guided exposure so they can learn what AI is good for, what it gets wrong, and how to use it responsibly in legal contexts.

Chilton frames the underlying issue as educators having been slow to adapt. He said educators have been “a little bit asleep at the wheel” by continuing to assign take-home work that students can complete with AI without the deeper learning the assignment was intended to test. And he links the concern to reports of AI cheating at Brown, Harvard, and other schools. When multiple institutions are seeing similar problems, the issue stops looking like a one-off “student ethics” story and becomes a structural assessment design challenge.

This is where the stakes move beyond one law school. In higher education, boards and leadership teams are watching for three things at once: academic integrity, student experience, and legal and reputational risk. If AI can reliably generate plausible text on demand, then traditional assessment formats stop being valid proxies for learning. That forces leaders to reconsider what they are actually testing and whether the mechanism still matches the outcome they want.

For decision-makers, the second-order implication is blunt: the “analog turn” in classrooms may become a template for how schools prove rigor while still preparing students for an AI-saturated workplace. The risk is that institutions keep trying to solve the problem with surveillance alone, or they swing too far and treat AI like a contaminant rather than a tool. UChicago’s approach is an attempt at balance: enforce offline thinking during key moments, then formalize AI literacy and tool usage so graduates are not entering practice unprepared.

For peers in other professional disciplines, the lesson is operational. Assessment design is now a product decision. If an exam can be cheated with an off-the-shelf model, it is not just a proctoring issue. It’s a curriculum integrity issue. UChicago is betting that redesigning the “how” of assessment, and pairing it with explicit AI instruction, is the fastest path to restore the connection between student effort and measurable understanding.