Berkeley Professor Stops Grading Student Homework

UC Berkeley math professor Richard Borcherds is dropping graded homework in favor of proctored, no-device exams.

Updated on Oct. 31, 2026 in Homework

Flat opaque gouache-painted illustration of a single fountain pen on a blank surface, evoking a shift in university academic policy. — UC Berkeley math professor Richard Borcherds has eliminated graded homework, opting for proctored exams to ensure student assessment remains free from AI influence. AI Illustration. Upload story photo >

Live Poll

Do you believe traditional grading can effectively measure student learning when AI tools are widely available?

University of California, Berkeley professor Richard Borcherds has stopped grading student homework assignments. The shift comes as the rise of artificial intelligence makes it difficult to verify if work is human-generated.

Why it matters

The decision addresses the challenge of distinguishing between student efforts and AI-generated content. Borcherds noted that traditional grading for correctness often encourages students to rely on AI tools rather than their own comprehension.

University of California, Berkeley policy allows professors to set their own AI guidelines after finding that available detection software is unreliable. Borcherds now relies on in-person exams that strictly forbid all electronic devices.

The players

Richard Borcherds

He is a professor at the University of California, Berkeley who has moved away from grading homework due to the prevalence of AI.

University of California, Berkeley

This institution allows individual faculty members to set their own policies on AI use within their course syllabi.

The details

Professor Borcherds now uses in-person exams to evaluate student knowledge, ensuring that assessment is not compromised by remote AI assistance. The University of California, Berkeley grants faculty the authority to determine individual classroom policies regarding AI usage.

October 31, 2026: Professor Borcherds' updated grading policy was reported.

Culture Shift

This policy shift mirrors a broader societal struggle to adapt education to generative AI tools, moving away from legacy homework models toward verified in-person assessment. It highlights the rejection of ineffective AI detection software in favor of traditional, device-free evaluation methods.

Students at the University of California, Berkeley should prepare for a transition to exam-heavy grading where performance is measured exclusively in the classroom. This change eliminates the reliance on take-home assignments, placing a higher premium on proctored, device-free assessment.

The takeaway

This development serves as a prime example of how educators are reclaiming the classroom environment in the face of rapid technological disruption. Students should focus on developing foundational skills that can be demonstrated without the aid of digital tools.