The University of Chicago is extending restrictions on AI and electronic devices from undergraduate social science courses to humanities classes, as the university seeks to strengthen students' independent thinking while debating how AI should be incorporated into higher education.
The restrictions, which took effect in the autumn quarter starting Sept. 28, require students to read printed materials and participate in in-person discussions without laptops, phones, smartwatches, or wearable devices capable of recording.
 |
|
An aerial view shows the Gothic-style campus of the University of Chicago, with ivy-covered buildings and autumn foliage lining the main quadrangles in Chicago, Illinois, the U.S., Aug. 17, 2026. Photo from the university's Facebook page |
Students are also prohibited from using AI to complete assignments, except for specified tasks such as dataset analysis or citation searches, which require instructors' approval.
The rules are intended to encourage students to develop their reading, writing, and analytical skills without relying on AI.
Professors are also restricted from using large language models to assist with grading unless they have conducted a thorough assessment and disclosed their use to students.
The University of Chicago currently ranks sixth in the United States according to U.S. News & World Report.
The expansion in AI restrictions comes as the university adopts a broader approach to AI that combines restrictions in some introductory courses with access to the technology elsewhere.
In June, the university announced a partnership with AI company Anthropic to provide faculty, staff, and students with access to its Claude tools.
Students must complete a Canvas module on responsible generative AI use before creating an account, according to The Chicago Maroon.
University President Paul Alivisatos has argued that students should learn to think both with and without machines.
"The first goal is that our students should learn during their time with us to be ethical, skeptical and ambitious about the role of AI in their education and their lives. The second is that all students at UChicago should learn to think both with and without machines," he wrote in The Wall Street Journal.
In an Oct. 12 article published by the University of Chicago Law School, William H.J. Hubbard, a law professor who chairs the university's faculty committee on AI and academic life, explained the rationale behind the contrasting approaches.
The university's law school introduced its own restrictions in July, prohibiting electronic devices in required first-year classrooms as part of a broader AI strategy.
The school plans to introduce AI more flexibly in higher-level courses, while requiring students to learn how to evaluate and improve AI-generated work.
"In a typical workplace setting, a tool that allows you to complete a routine task in five minutes rather than an hour becomes a business imperative. But in an educational setting, AI's superpower is educational kryptonite," he wrote. "In school, the goal is not output, and the measure of quality is not speed. The goal is learning, and the essential ingredients for learning are time and effort."
He added that this dynamic leads to the "AI learning paradox," where the way AI should be used in education is the exact opposite of how it is used professionally.
"If AI is going to be used for teaching and learning, it should help students devote more effort and attention, not less, to their work—by making studying more meaningful, entertaining, or challenging," he wrote.
Hubbard said the university’s restrictions had received positive responses but also attracted criticism.
The aim, he explained, was not to reject AI but to ensure students develop the judgment and reasoning skills needed to use it responsibly.
No university-wide AI policy
Despite the expansion of device-free teaching, the University of Chicago has not adopted a single AI policy covering all courses.
A report by The Chicago Maroon, based on interviews with 10 professors across disciplines, found that instructors were taking different approaches to AI use.
Some had redesigned assessments to require in-class writing or oral examinations, while others continued to allow AI for certain assignments.
The lack of a uniform policy can make expectations harder for students to navigate.
Jessie Wang, president of the university’s Student Advocate’s Office, told the newspaper that what constitutes misconduct can vary between courses, making it important for students to understand the rules set out in each syllabus.
The university’s official academic integrity guidance similarly advises students to check course-specific requirements and consult instructors when uncertain about whether AI tools or editing software are permitted.
The restrictions also come with practical considerations. Students in affected courses must adapt to paper-based reading and device-free classroom discussions, while instructors must design assignments and teaching methods around those conditions.
Exceptions may be available for students who require accommodations through Student Disability Services.
However, published evidence on the policy’s effects remains limited. As of Oct. 16, no systematic follow-up measuring changes in student engagement, academic performance, or learning outcomes since the autumn-quarter rollout had been identified.