A Brown University economics professor says he's uncovered what appears to be the largest AI-assisted cheating scandal in Ivy League history, after switching a midterm exam to take-home format for compassionate reasons following a campus tragedy. Roberto Serrano, who has taught economics at Brown for over three decades, gave students a take-home midterm this spring after many expressed anxiety about being in a classroom following a December mass shooting on campus that killed two students and injured nine.
The results immediately raised red flags. Of 86 students who took the exam, 40 scored a perfect 100, and the class average reached 96, dramatically higher than the historical range of 65 to 80, despite Serrano intentionally making the exam harder than usual since students would have unlimited time at home.
The Numbers Tell the Story
With his dean's support, Serrano switched the final exam to an in-person, closed-book format and told students plainly he suspected many of them had used AI to cheat. He gave the class a chance to prove him wrong. They didn't. Twenty-seven students subsequently dropped the course, 22 of whom had scored perfect 100s on the take-home midterm. When the in-person final took place, the class average collapsed to 48.6%, the lowest in the course's history, with several students who scored 100 on the midterm dropping to scores in the teens or lower on the final.
Serrano submitted his findings to Brown's Standing Committee on the Academic Code in May and initially received no response. After going public in late June, the committee asked him to file individual complaints against each suspected student, including copies of their exams, a process Serrano called ridiculous given that AI-detection tools are known to produce both false positives and false negatives.
A Pattern Playing Out Across Higher Education
This isn't an isolated incident. Recent reporting indicates 57% of U.S. college students now report using AI tools in their coursework on a weekly basis, and separate surveys found nearly half of surveyed Harvard seniors admitted to cheating. Serrano has already changed his approach for the coming academic year. Weekly homework will carry zero weight toward final grades, and take-home exams are permanently gone from his courses.
Why This Matters for Business
I've spent four years advising organizations on responsible AI adoption, and this story is a useful mirror for any business grappling with the same core tension. AI genuinely can help people learn and work faster, but only when the underlying skill development still happens. Serrano put it well when he told students that pressing a button to have AI do the work makes the human involved irrelevant to the outcome.
For business leaders, this is a preview of a much bigger workforce problem. The generation entering the labor market over the next several years will have been shaped by exactly this tension between AI-assisted convenience and genuine skill development. Companies hiring recent graduates should expect wider variance in foundational critical thinking and problem-solving skills than in prior cohorts, and should build verification and mentorship structures accordingly rather than assuming a degree guarantees underlying competency.
