A Law School Pulls the Screens
On 10 July, the University of Chicago Law School announced that it will prohibit electronic devices, laptops, tablets and phones, across all sections of its nine core first year courses during the 2026 to 2027 academic year. Exams in those courses will be administered in class, without access to the internet, electronic files or apps. There are narrow carve outs: professors can designate a classroom scribe to take shared notes electronically, use devices for activities like in class polling, and the school will accommodate disabilities as the law requires.
It would be easy to file this under nostalgia, another institution longing for the pre laptop lecture hall. That reading misses the point. The device ban is one visible piece of a year long strategy the school calls Rethinking Legal Education in the AI Era, developed with a 15 member AI Advisory Council. William Hubbard, the professor chairing the effort, captured the urgency by insisting the profession is no longer talking future tense here. The school is reacting to a present in which students can offload analysis to a chatbot faster than a professor can pose the next question.
The Three Pillars Behind the Ban
The strategy rests on three stated pillars. The first is developing AI resilient pedagogy and assessment, meaning course designs that reward sustained, effortful engagement and make shortcutting through AI less attractive. The second is elevating the essential human skills that distinguish excellent lawyers, the judgment, oral advocacy and client sense that no model supplies. The third is teaching the responsible, effective and ethical use of AI, so graduates leave fluent in tools they will use every day in practice. The device ban serves the first pillar without contradicting the third.
What makes the framing credible is that it refuses the easy binary. Dean Adam Chilton tied the strategy to the school's long standing goal of producing graduates prepared to be excellent lawyers, and David Gordon, a Sidley partner on the AI Council, said the school keeps rigorous analysis and the development of judgment at the heart of its curriculum, augmented but not replaced by AI. The institution is not claiming AI is a fad to be resisted. It is claiming that some foundational capacities have to be built the hard way before AI can safely augment them.
Assessment Is the Real Target
Strip away the headline about laptops and the more consequential change is to assessment. The school is adding a requirement that students defend their upper level substantial research papers in person, discussing completed drafts with the supervising professor. Legal research and writing will treat writing without AI as the foundation, layering AI instruction on top so students learn to supervise and critique machine output rather than depend on it. These are not gimmicks. They are attempts to make authorship legible again.
This is the part every education and corporate learning leader should internalize. When work products can be generated instantly, the take home essay and the unproctored exam stop measuring the learner and start measuring the tool. UChicago's answer is to move the point of evaluation to moments a model cannot sit in for the student: a live oral defense, an in class closed environment exam, a Socratic exchange. The device ban is not really about devices. It is a forcing function to relocate assessment to where human capability is actually observable.
Where AI Still Gets a Seat
Crucially, this is not prohibition dressed as strategy. The school is deliberately preserving spaces where students learn to use AI well. Clinics, where students represent real clients under supervision, are being set up as primary sites for hands on AI training, with each clinic developing tailored policies for responsible use. The stated goal is that students work both with AI and without it, so they graduate knowing when a tool sharpens their work and when it quietly corrodes it. Even in the restricted first year courses, AI uses that deepen learning, like clarifying background concepts before class or generating practice problems, are encouraged outside the room.
That balance is what separates this from the reflexive bans we have seen elsewhere. A blanket prohibition would leave graduates unprepared for a profession that has already absorbed these tools into daily practice. A free for all would let them skip the reasoning that makes them worth hiring. UChicago is trying to sequence the two, foundation first, augmentation second, and to make the sequence explicit rather than leaving each professor to improvise. For an institution, that coordination across nine courses is harder to pull off than any single rule, and it is the real work here.
A Template Other Faculties Will Study
We expect this to travel well beyond law. Every discipline that grades written analysis faces the same crisis of authorship, and most have responded with detection tools that do not work or honor codes that cannot be enforced. UChicago has instead redesigned the machinery of teaching and testing so the question of whether a student used AI matters less, because the assessment happens where a model cannot substitute for the person. That is a more durable response than any plagiarism scanner, and it is one deans in other faculties can copy.
Corporate learning leaders should notice too. The same logic applies to certification and internal training, where a completion badge earned by prompting a model proves nothing about capability. The organizations that get workforce AI training right will likely borrow UChicago's move: build foundational skill in controlled conditions, verify it through live demonstration, then teach augmented practice on top. The specifics differ, but the principle is portable. When the artifact can be faked, you assess the human directly, and you design the program so that assessment is unavoidable rather than optional.
The Tension Institutions Cannot Avoid
None of this is free of risk. A no device policy in 2026 will strike some students as regressive, and there are real accessibility and equity questions that the school's carve outs only partly answer. There is also a legitimate worry that graduates trained under heavy restriction could arrive at firms less fluent in the AI workflows those firms now expect, if the clinical and elective augmentation does not land as intended. The strategy's success depends on execution across dozens of independent minded faculty, which is never guaranteed.
Still, the honesty of the approach is what stands out. UChicago is not pretending it can detect its way out of the problem, nor that it can let students self regulate, nor that it can ban the technology its graduates will use for the rest of their careers. It is making a deliberate bet that some human capacities must be built before they are augmented, and it is willing to accept friction to protect them. Whether the specific rules survive contact with a full academic year is an open question. The willingness to redesign rather than police is the part worth emulating.
