96 on the Midterm, 48 on the Final: What Brown's AI Cheating Scandal Means for Your College Future
Imagine you’re a professor. A gunman just attacked your campus, killing two students — one of whom sat in your office days earlier, brimming with plans for her future. In that grief, you decide to show your students some grace: you’ll let them take their midterm at home instead of in a tense classroom.
Now imagine grading those exams and seeing something that makes your stomach drop. Forty out of 86 students scored a perfect 100. The class average is 96 — on an exam that was deliberately harder than any you’ve given before. In 34 years of teaching, the average had never topped 80.
That’s what happened to Roberto Serrano, the Harrison S. Kravis University Professor of Economics at Brown University, this past spring. What followed has become the largest AI cheating scandal in Ivy League history — and it’s a warning shot for every student, parent, and university in America.
The Anatomy of a Trap
Serrano isn’t just any professor. He’s a game theory specialist — the kind of person who literally writes the textbooks on strategic decision-making. (His textbook is the one Brown’s own economics department uses.) So when he saw those impossible scores, he didn’t panic. He strategized.
He ran his exam questions through ChatGPT. The AI produced answers that were technically correct but written in a distinctive, convoluted style — using contradiction arguments where simple direct proofs would do, for example. His graders found that same telltale pattern across dozens of student submissions.
But Serrano didn’t void the midterm. Instead, he sent his students a message that reads like something out of a game theory textbook:
“I am not declaring the midterm void for now. I am going to give the class a chance to prove me wrong. That is, if the distribution of the final exam is roughly similar to the distribution of the midterm, I will count the midterm. Otherwise, which is of course what I expect to happen, I will declare the midterm void and reweigh the final accordingly.”
The silence from his students was deafening. And telling.
The Collapse
When the in-person, handwritten final exam arrived, the numbers told the whole story:
- 18 students dropped the course before the final
- 9 more stayed enrolled but never showed up to the exam
- Of those 27 who fled, 22 had scored a perfect 100 on the midterm
- Among the 59 who actually sat for the final, the average score was 48 out of 100
- Three students scored zero
- The previous lowest final exam average in the course’s history had been 65
In the end, 19 students failed. Serrano voided the midterm entirely and made the final worth 80% of the course grade.
The professor’s summary was blunt: “The empirical evidence of fraud is overwhelming.”
Why This Should Scare You (Even If You’d Never Cheat)
Here’s the thing — this story isn’t really about cheaters getting caught. It’s about what happens to everyone when trust breaks down in education.
Take-home exams are dying. Serrano has announced he will never give a take-home exam again. He’s also removing weekly problem sets from his grading because they can be completed by AI. And he’s not alone. Princeton just ended its 133-year tradition of unproctored exams — a tradition where professors would leave the room during finals because students had pledged not to cheat under an Honor Code. That tradition, which survived two World Wars, the Cold War, and the internet, lasted exactly one semester into the ChatGPT era before Princeton faculty voted to require proctors in every exam room starting July 1, 2026.
The classroom dynamic is changing. When professors can’t trust that homework is student work, when they can’t give take-home exams without expecting fraud, the entire relationship shifts. Expect more in-class writing, more oral exams, more “explain your reasoning on the spot” moments. If you’re a student who actually does your own work — and most of you do — you’ll be navigating a world shaped by the actions of those who don’t.
Your degree’s credibility is at stake. This is the part nobody talks about. When employers and graduate schools hear that Ivy League students are scoring 96 on take-home exams and 48 in person, what do you think that does to the value of a Brown degree? Or a Princeton degree? Or any degree from a school that hasn’t figured out how to handle AI? Serrano put it starkly: “If we no longer defend truth and decency and honesty, then what kind of credibility are we going to have as academics?”
The Bigger Picture: AI Cheating Is Everywhere
Brown isn’t an isolated case. It’s just the one that got caught in the most dramatic fashion.
A 2026 survey of Princeton students found that 29.9% admitted to using AI to cheat on at least one exam or assignment. And that’s the admitted rate — the real number is almost certainly higher. Theo Baker, a Stanford graduate writing in The New York Times, was even more blunt: “I don’t know a single person who hasn’t used AI to get through some assignment in college.”
At Brown itself, the numbers are staggering. A university-wide report released the same week as the scandal found that 56% of undergraduates and 67% of graduate students use generative AI tools daily or weekly. Over half of Brown’s 3,000 audited syllabi had no policy whatsoever on AI usage.
The cheating tools are getting better, and the detection tools aren’t keeping up. AI detectors like GPTZero hover around 60% accuracy — barely better than a coin flip when the stakes involve expulsion. Studies show that with even minimal human editing, 88% of AI-generated content evades detection entirely.
What Brown’s Response Tells Us About Institutional Will
Here’s where the story gets uncomfortable for anyone who thinks universities will solve this problem quickly.
When Serrano submitted his evidence to Brown’s Standing Committee on the Academic Code in May 2026, he received no response. When he went public a month later, the committee asked him to file individual complaints against each of the 50+ suspected students — complete with copies of their exams — so they could run them through AI detection tools.
“They don’t get rewarded for dealing with a 60-person case of cheating,” said Tricia Bertram Gallant, director of the Academic Integrity Office at UC San Diego. “They are not incentivized to really prevent cheating or to report it when it occurs.”
Brown’s provost maintained complete silence throughout. The university’s academic code still doesn’t explicitly define what constitutes impermissible AI use. A spokesperson’s statement pointed to “formal adjudication” processes without addressing the systemic failure.
Meanwhile, Brown’s own GAITL Committee report recommended that faculty “de-emphasize punishment” around AI misuse and acknowledged that “norms are likely to change in the years to come.”
Serrano isn’t having it: “That cannot be the university’s position before an incident of this magnitude.”
What This Means for Applicants and Families
If you’re a high school student applying to college — or a parent helping guide that process — here’s what you need to take away from this:
The schools are scrambling. Elite universities don’t have this figured out yet. Brown’s academic code doesn’t mention AI. Princeton just gutted a 133-year tradition as a stopgap. When you’re evaluating schools, ask about their AI policies. A school with clear, enforced standards is a school that’s taking your education seriously.
Your genuine abilities matter more than ever. In a world where everyone can generate competent work with a chatbot, the ability to think independently, reason through problems in real time, and articulate ideas without technological assistance becomes more valuable, not less. The students who struggled on Serrano’s in-person final were revealing what happens when you outsource your learning.
Build skills that AI can’t fake. The students who did well on both the take-home and the in-person exam didn’t need ChatGPT. They actually learned the material. If you’re serious about thriving at an elite school — not just getting in, but actually succeeding there — invest in deep understanding rather than surface-level performance.
Honor codes are being rewritten in real time. The AI cheating wave is forcing every elite school to reconsider its academic integrity framework. By the time you arrive on campus, the rules will look very different from what they looked like two years ago. Be ready to adapt.
The Uncomfortable Truth
Roberto Serrano went blind at 17 from retinal dystrophy. He learned Braille, got into Harvard on the strength of his mind, and spent 34 years at Brown earning one of the university’s most prestigious appointments. When he describes his blindness, he frames it the way a game theorist would: “I view my disease simply as one more restriction that I have to deal with, and I optimize based on that.”
He didn’t let a profound physical limitation stop him from developing his mind. So you can imagine how it feels to watch students at one of the world’s best universities refuse to develop theirs — not because they can’t, but because a chatbot makes it easy not to.
“We cannot choose to become idiots,” Serrano told Inside Higher Ed.
He’s right. And the students who understand that — who actually want to learn, who see college as more than a credential to be gamed — will be the ones who thrive in whatever comes next.
Key Takeaways
- Take-home exams are effectively dead at elite universities. Expect more proctored, in-person, handwritten tests.
- AI detection tools are unreliable (~60% accuracy). Schools know this and are moving toward assessment methods that make cheating irrelevant rather than catchable.
- Nearly 30% of Ivy League students admit to AI cheating. The real number is likely higher. This is reshaping the college experience for everyone.
- Schools are not enforcing consequences consistently. Brown’s response was “meek” despite overwhelming evidence. When evaluating colleges, look at how they handle integrity, not just what their honor code says.
- The students who actually learn the material will have a massive advantage. As AI makes surface-level performance cheaper, deep competence becomes rarer and more valuable.
Sources: Inside Higher Ed (Jul 8, 2026); Ars Technica (Jul 9, 2026); El País English (Jun 28, 2026); Fortune (Jun 29, 2026); Futurism (Jun 2026); The Brown Daily Herald (Jul 10, 2026); Brown GAITL Committee Report (Jul 7, 2026); The Atlantic (May 2026); Princeton Alumni Weekly.