Background & Context§
Artificial intelligence has infiltrated every corner of academia, and elite universities—long bastions of prestige and trust—are now grappling with its consequences. The latest incident at Brown University underscores a growing crisis. Professor Roberto Serrano, a 61-year-old blind economist and Harrison S. Kravis University Professor, discovered a massive fraud in his advanced mathematical economics course, ECON 1170. The midterm exam, administered on March 5, 2026, yielded an average score of 96 out of 100, with 40 students achieving perfect scores. Suspicions arose when graders noticed unusual passages matching ChatGPT outputs. This case is believed to be the largest known cheating scandal in Brown's history and across the Ivy League.
The News: What Happened Exactly§
Professor Serrano, a blind economist who has taught at Brown for 34 years, decided to administer a take-home, closed-book midterm exam for ECON 1170 in spring 2026. The decision was partly influenced by a tragic campus shooting in December 2025—a former PhD student opened fire during a review session for Introduction to Economics, killing two and injuring nine. Two of Serrano's own students were among the injured. To ease student anxiety, Serrano opted for take-home exams.
The midterm results were extraordinary: the average score of 96/100 and 40 perfect scores immediately raised red flags. Graders flagged answers containing passages later identified as ChatGPT outputs. Serrano did not void the midterm but warned that the final exam (worth 50% of the grade) would be held in person and that if grade distributions differed significantly, only the final would count. The final exam average plummeted to 48/100. Of 89 students who took the midterm, only 59 sat for the final. Notably, 22 of the 27 absent students had scored 100 on the midterm.
Serrano presented conclusive evidence of fraud—at least 50 students—to university officials. The president remained silent, and the dean did not comment until the case reached the Academic Code Committee, which labeled it a "wake-up call." Serrano expressed dismay: "That cannot be the university's position before an incident of this magnitude. Academic integrity is a value worth defending." He argues that wealthy families' donations often shield students from consequences, making faculty feel isolated in defending academic standards.
Historical Parallels & Similar Incidents§
This scandal echoes a broader collapse of trust in take-home exams sparked by AI. In 2024, Princeton University ended a 133-year-old honor system that allowed unproctored exams. Since 1893, Princeton had relied on an Honor Code where professors left the room during exams and students policed themselves. AI rendered this system obsolete, as cheating became "easier and more remunerative than ever," as Theo Baker wrote in The New York Times. Baker, a Stanford graduate, noted, "I don't know a single person who hasn't used AI to get through some assignment in college."
The parallel is striking: both Princeton and Brown faced the same core issue—AI enabling large-scale cheating in environments previously governed by trust. Princeton's solution—mandating in-person proctoring—mirrors Serrano's planned reforms: eliminating take-home exams and reducing online exercises. However, Brown's case goes further by exposing administrative inertia. While Princeton acted preemptively, Brown's leadership hesitated, potentially incentivizing more fraud.
Another historical parallel lies in the early 2000s plagiarism scandals triggered by websites like Chegg and Course Hero. In 2020, the University of Michigan discovered widespread cheating using Chegg during remote exams, leading to increased proctoring software use. Yet AI is more insidious because it generates unique, original-sounding answers that bypass traditional plagiarism detectors. Serrano's detection relied on pattern recognition of ChatGPT-style prose, a method akin to how teachers previously identified copied text from online "homework help" sites.
What sets Brown's case apart is its raw empirical evidence: the dramatic grade collapse from midterm (average 96) to final (average 48) and the selective absence of high-scoring students. This data-driven proof eliminates ambiguity. The incident also underscores AI's role as an equalizer in cheating—lowering barriers for students who might otherwise lack resources for human ghostwriters.
Finally, the shooting at Brown adds a tragic dimension. Serrano's take-home policy was a compassionate gesture, yet it was exploited. This juxtaposition highlights how AI can exacerbate already vulnerable situations. The lesson: universities must anticipate that AI makes trust-based systems fragile, and proactive measures—not reactive silence—are essential to preserve academic integrity. As Serrano poignantly asks, "If we no longer defend truth and decency and honesty, then what kind of credibility are we going to have as academics?"