Mass Use of ChatGPT to Pass Exams: Overwhelming Evidence Against 50 Students

The economics professor Roberto Serrano at Brown University has uncovered a mass fraud involving at least 50 students in his advanced mathematical economics course, ECON 1170. The overwhelming evidence points to a widespread use of artificial intelligence to cheat on the midterm exam in March, marking the largest scandal known not only at Brown but in the entire Ivy League.

The initial reaction from university leadership was one of coldness and silence, both from the president and the dean, who acknowledged the incident as a "wake-up call" only after Professor Serrano's intervention to the Academic Code Committee.
"This cannot be the position of the university in the face of an incident of such magnitude. Academic integrity is a value to defend. The faculty cannot be left alone in a decisive battle for the future of higher education," stated Serrano, a 61-year-old economist from Madrid who has been teaching at Brown University for 34 years.

The course, a high-profile offering that Serrano has taught for years, traditionally attracted few but excellent students, with classes rarely exceeding 30 enrollees. However, this semester, likely due to a new assessment system, 86 students enrolled. For the midterm exam, held on March 5, Serrano opted for a take-home, open-book test with practically unlimited time, allowing him to formulate more complex questions, often altering the assumptions of the model discussed in class.

The results of the test were "extraordinary": a 96 out of 100 average, with 40 students achieving the highest score. However, those who graded the exams immediately reported several anomalies.
"Some answers contained unusual passages that matched the output from entering the questions into ChatGPT," explained Serrano, adding substantive elements to his case.

The professor did not cancel the midterm exam but warned students that the final exam, which counts for 50% of the overall grade, would be held in person. He also clarified that if the distribution of grades was not similar to that of the first exam, only the final exam score would be considered. The results spoke clearly: the average plummeted to 48 out of 100. Of the 89 students who took the midterm exam, only 59 showed up for the final, and among the 27 absentees, a staggering 22 had received a perfect 100 on the first exam.
"The empirical evidence of fraud is overwhelming," concluded Serrano.

For the upcoming academic year, the professor has already decided to implement structural changes: weekly exercises will no longer count toward the final grade, given the risk of AI use. Additionally, take-home exams will be abolished, although in some circumstances they may be considered a necessary compromise.

The decision to organize take-home exams follows an event from last December 13, when Brown University made headlines due to a tragic event: a shooting on campus that caused two deaths and nine injuries, two of whom were enrolled in Serrano's course. Deeply affected, the professor had thought about offering relief to students, many of whom were still shaken by the anxiety of returning to campus.

The issue of artificial intelligence is altering centuries-old traditions at elite U.S. universities. Princeton, for example, has suspended a practice that lasted 133 years: since 1893, exams were unsupervised due to an Honor Code that required students not to cheat. Now, professors must supervise in-person exams, prompted by an urgent necessity.
"AI has made cheating easier and more rewarding than ever," wrote journalist Theo Baker in another article for the New York Times, adding that he does not know anyone who hasn't used AI to pass at least one university assignment.

Serrano agrees that AI increases the incentives to cheat and reiterates the necessity not to sweep these cases under the rug. On the contrary, they should help open an in-depth debate.
"If we no longer defend truth, decency, and honesty, what kind of credibility can we have as academics?" concluded the professor, highlighting the urgency of a constructive dialogue for the future of higher education.