Staff Reports
Brown University’s Roberto Serrano, the Harrison S. Kravis University Professor of Economics, blew the whistle this week after a take‑home midterm blew up into what he calls an “overwhelming” case of AI‑assisted cheating. A compassionate change to help students after a campus shooting exposed a larger problem: generative AI tools like ChatGPT are being used as shortcuts, and university leaders are not answering the bell.
Take‑home exam turned red flag
The numbers jump off the page. Of 86 students who took the closed‑book, take‑home midterm in ECON 1170, 40 scored a perfect 100 and the class average hit the mid‑90s — far above the usual 60s to 80s. Serrano and his graders found the same odd, convoluted reasoning in many papers. When he ran the questions through ChatGPT, the model produced matching passages. That pattern across dozens of submissions is not a coincidence. It is textbook AI cheating.
Professor acted — students reacted
Serrano did what a good teacher should: he called the class out and set a high‑stakes in‑person final to check the results. After his warning, 27 students dropped the course — 22 of them had perfect midterms. Only 59 showed up for the in‑person final and 19 failed. The final average cratered to the 40s. Serrano’s verdict: “The empirical evidence of fraud is overwhelming.” That is a hard claim to shrug off.
Administration silence is the real scandal
Here’s where the story goes from academic concern to institutional failure. Serrano says he reported the evidence to the Dean of the College, Rashid Zia, and to Provost Francis J. Doyle III. At first, he got no response. Only after he escalated did the Academic Code Committee call it a “wake‑up call.” Silence from the provost is not just tone‑deaf — it looks like a refusal to protect academic integrity. If Brown’s leadership won’t support faculty enforcing standards, the university risks turning its diploma into a price tag with no value attached.
A conservative demand: standards, not excuses
Generative AI is real and useful. But so are exams, learning, and honesty. Colleges must stop treating AI as an abstract policy memo and start treating it like classroom reality. That means clear investigation under Brown’s Academic Code, honest public answers from Provost Francis J. Doyle III and Dean Rashid Zia, more in‑person assessments, and tools to detect AI misuse. Parents and employers deserve graduates who can think, not students who can press buttons. If universities won’t act, voters and boards should.
Brown’s episode is a warning shot. This isn’t about policing technology for technology’s sake. It’s about defending learning, upholding academic integrity, and making sure an Ivy League name still means something. If administrators keep mum and hope problems vanish, the next tsunami won’t be generative AI — it will be the collapse of standards that took decades to build.
