A UC Berkeley professor has reportedly abandoned grading assignments after noticing that his students have been abusing AI to secure top scores.
“Student use of AI for homework is now widespread,” Richard Borcherds told the California Post. “Grading for correctness just adds a huge incentive for them to use AI for homework instead of doing it themselves, and penalizes students for honesty.”
Borcherds — a renowned mathematician who received the prestigious Fields Medal in 1998 — added that there is “no point” in paying to have students’ work evaluated for artificial intelligence, calling it a waste of time and money.
Instead, the professor has turned to giving his students credit for simply completing their work — paying no mind to whether the answers are right — and admits he knows of no better approach.
“There is no easy way to tell the difference between human and AI generated work,” Borcherds told the newspaper. “Occasionally one could tell it is AI because AI made a few characteristic errors, but AI has now improved so much that it rarely makes obvious mistakes.”
The professor also expressed his concerns about the future of mathematics in an interview with UC Berkeley’s student newspaper, the Daily Californian.
“If AI continues to improve at present rate, then, in a few years, there’ll be nothing human mathematicians can do that AI can’t do better,” Borcherds warned.
“I have no idea what to advise,” he added. “I even asked ChatGPT this exact question, and even ChatGPT said it didn’t have any idea what was going to happen.”
The professor went on to say that he no longer sure whether to advise students to major in mathematics, noting that five years ago, a mathematician appeared to be one of the safest careers from the AI boom.
Borcherds compared the situation to physical paper journals, which are still published, despite being “utterly pointless” for two decades.
“There have been plenty of technical breakthroughs in the past — hand weaving was put out of business by looms, and horses were put out of business by motor cars, and so on,” he told the Daily Californian. “This is the first time anything like this has happened in mathematics, though.”
The university has also tested AI-detection tools, such as Turnitin and Rumi, but concluded they are not adequate enough to implement as a standardized campus-wide solution — leaving professors to determine how they want to handle to issue of AI in their classrooms.
“If the worst effect of AI on society is that it is harder to assess homework then I will be very relieved,” Borcherds told the Post, suggesting he anticipates larger problems ahead.
Breitbart News previously reported on the “Cobra Effect” of AI detection tools, which are causing students to write simpler and less descriptive essays:
Writing instructors and students across educational institutions are discovering an unintended consequence of AI detection tools: rather than simply identifying machine-generated text, these systems are fundamentally reshaping student writing practices in concerning ways. The tools, which rely on statistical analysis rather than genuine comprehension of authorship, are penalizing sophisticated writing while rewarding generic, algorithm-friendly prose.
A striking example emerged when a student’s essay about Kurt Vonnegut’s Harrison Bergeron received an eighteen percent AI-generated score from a detector pre-installed on a school-issued Chromebook. The trigger was a single word: “devoid.” When the student replaced this term with the simpler word “without,” the AI detection score dropped to zero, despite no changes to the essay’s underlying structure or ideas. This incident illustrates how current detection systems function on surface-level statistical signals such as word choice and distribution rather than meaningful assessment of authorship.
The technical foundation of these detectors contributes to the problem. AI detection tools estimate the probability that text was machine-generated by analyzing features including token frequency, syntactic patterns, and burstiness, which refers to variation in sentence length and structure. Students have learned that certain stylistic markers can trigger higher detection scores, while flatter, more generic text tends to pass unnoticed. This creates a rational but problematic response: students either deliberately simplify their writing or use AI models to generate statistically safe phrasing that blends into the background.
Breitbart News social media director Wynton Hall, author of the New York Times Bestseller Code Red: The Left, the Right, China, and the Race to Control AI, understands the danger of unintended consequences of advanced technology. Hall explains, “America has one foot in the roses of possibility and promise of technological innovation, and one foot hovering over the landmines that we are going to have to collectively navigate our way though as a society. Every one of those landmines touches every policy specialty that the conservative movement has entered.”
Alana Mastrangelo-Wrask is a reporter for Breitbart News. You can follow her on Facebook and X at @ARmastrangelo, and on Instagram.
COMMENTS
Please let us know if you're having issues with commenting.