California professors are responding to AI in markedly different ways, from paper exams and completion-based homework grading to using chatbots as teaching aids. The practical shift is already visible: some instructors are moving assessment into the classroom, where students cannot quietly hand the work to a model.
Watch Desk analysis
What happened
The New York Post interviewed professors at UCLA, USC, UC Berkeley and Stanford about AI in their teaching. UCLA lecturer Thomas Maierhofer says his exams are in person, on paper and without phones or laptops, while homework and labs are graded for completion. He is also developing a chatbot intended to support student learning.
At Berkeley, maths professor Richard Borcherds says he now bases grades mainly on in-person exams without electronic devices. USC education professor Morgan Polikoff says he had already planned to move more graded work to sit-down paper tests. Stanford Law professor Julian Nyarko takes a more optimistic view of AI as a teaching tool, while warning against using it to produce work in place of learning.
Why it matters
There is no single campus-wide answer here. The examples show instructors changing what they assess, where they assess it and what they expect AI to do. That matters to students facing different rules from one course to the next, and to universities trying to decide whether AI belongs in the classroom or mainly outside it.
The reporting also puts a useful distinction on the table: using AI to help explain a subject is not the same as submitting its answer as your own. That distinction sounds tidy. Designing assessments that can tell the difference is rather less so.
Our read
The most useful response may not be a blanket ban or a cheerful instruction to embrace the bot. These professors are testing a more specific question: what should an assignment measure when a student can ask AI to produce a plausible answer? In-person exams and completion-based homework are clear choices, but they also change what counts toward a grade. Students should check each instructor’s rules rather than assume one university-wide policy settles the matter.
What to watch
- Whether more courses move important assessments into supervised, device-free settings.
- Whether completion-based homework is paired with other ways to demonstrate understanding.
- How universities distinguish permitted learning support from work submitted as a student’s own.
Discussion spark: Should universities move more assessment into supervised, device-free exams, or is that too high a price for making coursework harder to outsource to AI?
Sources and evidence
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.