Last March, professor Roberto Serrano, a tenured professor of Economics at the prestigious Brown University (United States), got a major surprise while grading his mathematical economics exams. Of the 86 enrolled students, at least 50 had used artificial intelligence to answer the questions. And it wasn’t just a suspicion: the professor had compelling evidence.
Serrano, who is also blind and uses AI as a support tool in his daily life, designed a take-home exam but without outside help. When he saw the results were extraordinarily good —and suspiciously uniform— he applied an AI detector and confirmed the massive fraud. What is most worrying is that, when he reported the case to the university’s administration, their first response was silence. «We need to publicly admit the seriousness of the situation and open a deep debate,» he told the newspaper El País.
This case is not isolated. More and more universities around the world face the same dilemma: how to evaluate students when AI is capable of acing university exams? The solution is not only about banning the technology, but about rethinking the educational system. Meanwhile, if you’re a student, the moral is clear: artificial intelligence is a great tool for learning, but using it to cheat can be costly. And it’s getting harder and harder for it to go unnoticed.
— Marta, for intermittent intelligence.
