A Brown University economics professor's grade book just proved what working Americans already suspected: elite university credentials are increasingly worthless, and AI is making it impossible to hide. When Professor Roberto Serrano switched from a take-home midterm to an in-person final, scores plummeted from an average of 96 percent to roughly 48 percent, with 19 students failing outright — after nearly half the class had scored perfect or near-perfect on the take-home exam they almost certainly cheated on using AI.
The stakes are straightforward. Parents spend hundreds of thousands of dollars and students take on crushing debt for Ivy League degrees that are supposed to signal competence. If those degrees signal nothing more than the ability to prompt ChatGPT, the entire credentialing racket collapses — and the people who pay for it get stuck with the bill.
Serrano teaches Welfare Economics and Social Choice Theory. He converted his midterm to a take-home format after a December campus shooting killed two students, an accommodation meant to reduce stress. Instead, 40 of his 86 students scored perfect or near-perfect — compared to typical midterm averages between 65 and 80 percent. Serrano told Business Insider that he and his teaching assistants compared student responses with ChatGPT outputs and found similar wording and reasoning.
"The problem with this technology is that the cost of cheating has basically gone down to zero," Serrano told Business Insider. "It's very easy for students to succumb to the temptation."
When Serrano announced the final would be in-person — warning students that if the distribution diverged from the midterm, he would void the midterm results — several high-scoring students dropped the course. Among those who stayed, the results were devastating. Students who scored in the high 90s on the midterm scored in the 50s on the final. Of 59 students who took both exams, only two showed consistent performance: one student scored a 95.5 and then a 95, another scored a 55 and then a 59.
"Since I'm a big defender of integrity, yes, I would hire that person," Serrano said of the consistent high performer, whom he said he knew "very well."
The Daily Caller framed the scandal as AI "making students dumber." Business Insider and NewsBreak played it more cautiously, noting the grade distribution "isn't a perfect study" and that other factors could explain variance. That's technically true — but when 40 students suddenly ace a harder-than-usual exam and then collapse the moment they're in a room without internet access, Occam's Razor cuts one way.
The university's response tells its own story. Serrano submitted the data to Brown's Standing Committee on the Academic Code on July 8. The Daily Caller reported that the committee initially did nothing, and Serrano went public in June out of frustration — after which Brown officials requested formal academic integrity complaints be filed against each suspected student. Brown VP Brian E. Clark wrote that the university "treats every allegation of academic integrity with the utmost seriousness." It took a professor going public to get them to act.
The tech world noticed. Y Combinator cofounder Paul Graham posted the score chart on X, and two Google DeepMind staffers weighed in. The debate turned to workforce readiness: can students who cheat with AI be trusted to do real work? Serrano's answer was unequivocal — hire the person who showed up and performed honestly, not the one who gamed the system.
Serrano says he will never administer a take-home exam again and is eliminating the homework portion of his grades entirely. "It's certainly a wake-up call to the professors," he said. "We need to pay attention to this."
The open question is whether the rest of the academy will follow — or whether they'll keep pretending the credentials they sell still mean something.