Artificial intelligence is reshaping college classrooms, prompting some of California’s best-known professors to revive an decidedly low-tech safeguard: handwritten work on paper.
The California Post spoke with professors at UCLA, USC, UC Berkeley and Stanford about AI’s impact on their courses. Some said they are largely eliminating the technology and grading only written assignments completed in class without digital assistance.
Others, however, are incorporating AI into their teaching and view it as a potentially valuable addition to the classroom.
“The technology has been shoved down all our throats and we’re all just dealing with the fallout,” Morgan Polikoff, a professor at USC Rossier, told the Post.
Polikoff, who is currently on leave from teaching, said he had already intended to move more graded undergraduate work to supervised, paper-and-pencil exams.
He said he might adopt an approach like that of prominent UC Berkeley mathematics professor Richard Borcherds, who drew attention after deciding he would no longer grade homework based on correctness.
Two years ago, Polikoff caught a freshman using AI to write a paper despite a ban on the technology. He said the use was easy to identify, and the student admitted it when challenged. Polikoff then required him to start the paper over.
“I think in general we will need more standardized and AI-proof measures of everything, both in college admissions and once students are enrolled at university,” he said. “It’s become increasingly difficult to know what anyone knows and can do, and that signal is important.”
“I don’t see any point at all in a student completing an assessment by running it through AI and handing in to me whatever it spit out for them—that is a waste of my time and theirs,” he added.
USC leaves AI policy largely to individual instructors, allowing them to establish rules for their own courses. Polikoff said his PhD program is currently discussing a policy, although faculty would still retain discretion over how to apply it.
The university spends about $3.1 million annually on a ChatGPT Edu subscription for 80,000 users. Polikoff described that cost as “pennies” compared with USC’s payment to its football coach.
At UCLA, statistics and data science lecturer Thomas Maierhofer has adopted a middle-ground strategy.
“All exams are in-person on-paper exams without the use of phones or laptops. All work done outside of the classrooms like homework and labs are graded on completion,” he said.
Maierhofer also sees AI’s potential as an instructional resource. He is developing ALISSTAIR, a custom chatbot designed to assist students, while relying on his exam format to protect academic honesty.
“As long as the exams are in-person and on-paper students are only hurting themselves if they use AI to cheat instead of using it to enhance their own learning,” he said.
At UC Berkeley, Borcherds said his grades now depend primarily on in-person exams, with students prohibited from using electronic devices. His decision to stop grading homework for correctness made headlines this fall.
“Online exams are now a waste of time as several students just use AI to answer the questions,” he said. “I cannot tell which students are using AI, but I notice the average score increases a lot in online exams.”
DON’T MISS: Taylor Swift Shines at Nashville Songwriters Hall of Fame Induction
The University of California allows individual campuses and academic programs to establish tougher AI restrictions, particularly in areas such as admissions and academic integrity.
Stanford Law School professor Julian Nyarko takes a more optimistic view of AI—though he is less confident in people.
“I think AI has tremendous potential. I think many humans are not great teachers. So I’m skeptical of human capability, more so than I am overly enthusiastic about AI capabilities,” he told the Post.
“I think there’s great variance in the quality of teachers and I think AI can definitely have a positive impact if we learn to use it well, especially if the counterfactual is having a poor teacher,” he added.
In a study Nyarko led this year, 16 law professors blindly rated answers to common student questions, and they preferred Google’s AI over their own colleagues 75% of the time. The researchers stressed that the study measured answer quality, not whether AI tutors help students learn in the long run.
He also warned that there is a wrong way to use the tech.
“There’s a way to use AI that will harm students, an irresponsible use, which is basically, just write this paper for me or solve this question for me,” he said.
Nyarko has always given in-class exams.
He said many of his colleagues used to assign take-home exams lasting as long as eight hours. With one or two exceptions, they now feel they can’t, out of fear that students will simply hand the work to AI.
He said that trying to restrict students’ access is a fantasy at this point, since the vast majority already have it.
“AI will require significant change in how we do things,” he said. “We’re basically moving from a world where we kind of thought we knew how education works reasonably well, to a world where we don’t know how to best educate students. We have access to AI, and that requires a lot of figuring out.”
Stanford gives all students free access to ChatGPT, Gemini and Claude through a pilot running until August 2027. Unless an instructor sets a stricter rule, students may use AI to help them learn but can’t pass off its work as their own.
The debate has spilled into next week’s race for state schools chief.
Sonja Shaw, the Republican candidate for California Superintendent of Public Instruction, said over-reliance on AI robs students of the critical skills that colleges and employers demand. She said the fix is a return to academic rigor and objective standards.
“More AI-assisted work and grade inflation made it easier to disguise who actually learned the material and who didn’t,” Shaw said. “Objective testing remains the best means to measure student skills in an environment where they can’t simply rely on ChatGPT to do the work for them.”
“AI should, at best, be viewed like the calculator: it may eventually be useful as a tool, but can be dangerous if you don’t first understand the foundation of what it’s helping you accomplish,” she added.
