Concerns Over AI Cheating at Brown University
Roberto Serrano, an economics professor at Brown University, recently criticized the institution’s slow reaction to a significant cheating scandal involving artificial intelligence (AI). In an op-ed published in Free Press, he reflected on the university’s decision to transition midterm and final exams online following a mass shooting on campus in December, which led to a remarkable increase in course enrollment—from 30 to 86 students.
Serrano reported that his class’s average score on the online midterm surged to 96 points, sharply contrasting with the prior year’s range of 65 to 80 points. Alarmingly, 40 students scored perfectly, often with answers that closely mirrored those generated by ChatGPT.
Shifting Exam Formats
Due to these events, the university decided to revert final exams to a traditional in-person format. Additionally, it invalidated midterm scores that deviated significantly from final results. This change led to only 59 students sitting for the final exam, resulting in an average score of 48.6 points, with a max of 95 and a min of 0.
Interestingly, of the 27 students who opted out of the course and exam, 22 had previously achieved perfect midterm scores.
Serrano shared these findings with Brown’s Academic Regulations Committee and the campus paper, but he felt ignored until his concerns gained wider attention last month.
A Call for Integrity in Academia
“Academia is meant to be a beacon of truth,” Serrano expressed. “We must not tolerate low moral standards among promising students. What transpired in my class serves as a cautionary tale. If we overlook and fail to punish such actions, we may inadvertently reinforce dishonest behavior.”
In response to Serrano’s allegations, Brown University stated that they have “consistently engaged” with him regarding the issue of AI fraud. They indicated that academic leaders reached out to him in May 2026 to discuss formal measures for addressing his claims. After Serrano provided details to the Academic Standards Standing Committee on July 8, they assured him that the investigation would advance.
AI’s Role in Educational Cheating
Serrano’s worries extend beyond just his own classroom. He mentioned that if not for publicizing his experience, the university might not have addressed the issue at all. “I’m grateful for this progress, yet I truly believe none of it would have happened had I not gone public. Brown, like many universities, is wrestling with how to effectively incorporate AI into its academic framework,” he noted.
In his article, he rebuffed the notion that AI is a necessary tool for students under pressure. He argued that competitive challenges have always existed and should not excuse cheating. “During my midterms, I allowed nearly unlimited time for students. They weren’t desperate; they were intentionally cheating to score higher,” he remarked.
Serrano concluded by acknowledging that while the temptation to cheat will always exist, implementing the right incentives is essential for promoting better behavior among students.
