When Roberto Serrano, a professor at Brown University, changed the format of his midterm exam last spring, he had his students’ mental health in mind, not academic cheating. Two of them had been shot, including Ella Cook, a young woman who had been sitting in his office just days earlier December 13 massacre at Brown University and asked him to be her academic advisor.
“We had a very nice conversation,” Serrano recalled in an interview with Assets. “She was a wonderful young woman, full of energy, full of ideas. Imagine my shock when, a few days after that conversation, the names of the two fatal victims were released and I saw that one of them was her.”
In this grief, Serrano made a decision he had never made in his 34-year career at Brown and gave his ECON 1170 course — an advanced undergraduate course in mathematical economics — a midterm to take home. He wanted to eliminate the stress that comes with sitting in a classroom on a campus where, he says, many students were still too traumatized to set foot. Two of his students were among them the nine injured in the attack; They fought for their lives for weeks and both survived.
What Serrano got instead of gratitude was the largest known AI-powered cheating scandal in the Ivy League, as previously reported El Pais.
Fraud on a mass scale
Of the 86 students who took the exam on March 5, 40 received a perfect score of 100. The class average was 96, whereas in previous years the average had been between 65 and 80 – and this exam was inherently more difficult than usual. “The nice thing about take-home exams used to be that we professors could challenge students a little more just to push them to a higher level,” Serrano said. “The fact that this was a more difficult test and this distribution made it clear that something very unusual had happened.”
Serrano got a tip that was just too clever, he said. “Some answers contained unusual passages that were consistent with the results obtained after ChatGPT answered the questions,” Serrano said. His evaluators administered the exam questions via ChatGPT and made a telling discovery: the AI had generated a convoluted argument for a problem for which there is a much simpler and more elegant proof, and the same convoluted argument appeared in dozens of student exams. “This distribution made it clear that something seriously wrong had happened,” he said, calling it “absolutely ridiculous.”
But Seranno said he decided to give his students the benefit of the doubt: He wouldn’t cancel the midterm exam, but would tell them the final exam would be held in person. If the distribution of grades was not close to that of the midterm, only the final would count.
When Serrano returned to class after grading, he told his students exactly what he had found. “If you have done this, if you simply press a button to ask an AI agent to do this for you, you prove yourself to be completely irrelevant. So my question to you is: Why are you here? Why are you at a university if you refuse to learn, to work hard, if you refuse to put in the effort necessary to develop critical thinking?”
“If you just press a button to have this machine do the work for you, do you think you need a Brown degree to do that?”
When asked about his students’ initial reaction, Serrano responded with just one word: “Silence.” He suspected the cheaters weren’t even there: “I honestly think most of the cheaters weren’t in class.” He ended class that day by reminding students of the honor code. “You all signed that, right? Unfortunately, that’s the value of your signature.”
After his speech, 27 students dropped out of the course; 22 of them had scored 100 points at take-home.
When the final came around, only 59 appeared for the personal exam and 19 failed. The class average plummeted to 48 out of 100: by far the lowest final exam average in the course’s history. “The empirical evidence of fraud is overwhelming,” Serrano said. “When you put all this information together and the distributions of the two tests, it is absolutely clear.”
After compiling his evidence, Serrano sent it to Brown’s dean of the college and provost. Neither of them responded at first. After referring the case to the university’s Academic Code Committee, he received a message calling the incident a “wake-up call.” The provost, he said, had maintained complete silence until today.
The man who wrote the book on game theory explains game theory
Serrano holds a named chair, the Professor of Economics at Harrison S. Kravis Universityone of the most prestigious appointments a university bestows. He is an editor at Games and economic behaviorthe leading journal in a field concerned with the economics of risk, uncertainty, and information, often known as “game theory,” and exactly what is at play when, for example, you cheat on an exam.
Serrano has over 6,100 citations on Google Scholar and is the author of two widely used textbooks, including one Brown’s own economics department used. He is a member of renowned academic societies and even received the King of Spain Prize in Economics in 2024.
The game theory expert looks at the current situation and despairs. “I’m very frustrated,” Serrano said Assets. “I believe that the arrival of AI was like a tsunami for all of us. It caught everyone off guard. But in my humble opinion, silence is the worst solution to this problem.”
Serrano, who has been blind since age 17, earned his doctorate from Harvard and spent more than three decades at Brown University, acknowledged that AI had evolved so quickly that institutions didn’t know how to respond. Brown has not yet responded Fortune’s Please comment.
But it’s not just Brown, Serrano said. He pointed to a youngest New York Times Essay It described a pervasive culture of AI fraud among Stanford colleagues: students who attended elite universities not to learn but to collect credentials. “What they miss in this very naive analysis,” Serrano said, “is that the Brown label is Brown for a while. But if Brown continues to produce mediocre students who refuse to learn, sooner or later the market will find out that the Brown label is not what it once was.”
The wider trajectory, he warned, suggests something darker. “If workers simply push a button and ask an AI agent to do the work for them, that imprints a world in which humanity has chosen to become idiots,” he said. “We stop thinking.”
Brown is far from alone. Princeton faculty voted in May to end its 133-year-old Honor Code tradition of unproctored exams and require examiners in every room starting July 1, the most significant policy change since students first petitioned for them in 1893. As Assets reported in May57% of U.S. college students now report using AI tools in their coursework on a weekly basis. A separate one Assets The analysis showed this AI causes measurable cognitive atrophy among students, with educators warning of a “great unraveling” of the ability to think independently. And just last week, 47% of Harvard graduates surveyed agreed to cheat.
Serrano has already made changes for the upcoming academic year. Weekly homework has no impact on final grades as it can be completed using AI. There are no more take-home exams forever. “Unfortunately, the idea of a take-home exam is a thing of the past,” he said. “It’s too easy for students to succumb to temptation.”
“I’m sure there are useful uses for AI: it has the potential to be something very useful for students and contribute to learning,” he said. “But we must be absolutely clear about the risks this poses to academic integrity, and that is a value we cannot abandon.”
For Serrano, the last word is not about exams or grade distribution. It’s about what kind of people universities produce. “We must create the necessary guardrails – and if they fail, we must be prepared to draw consequences,” he said. “But this is bigger than science.” “If we no longer defend truth, decency and honesty,” Serrano said, “what kind of credibility will we have as academics?”