AI Cheating Scandal at Brown University: Professor Warns of a "Failed Society"
Professor Roberto Serrano of Brown University's Economics Department exposes the reality of AI-enabled cheating. Following the introduction of remote exams, the average score skyrocketed to 96, with nearly half of the students achieving perfect marks. Serrano warns, "Entrusting judgment to AI is choosing the path of becoming fools."
Students at Ivy League universities are expected to possess the intelligence necessary to understand textbooks and tackle exams. However, the pressures of competition, ambition, and packed schedules have driven some toward the “shortcut” offered by generative AI. This trend has now surfaced prominently at Brown University, where Economics Professor Roberto Serrano revealed alarming results after introducing remote exams in his course. According to Serrano, the average score in one of his classes soared to 96, with 40 out of 86 students achieving perfect scores. The gravity of the situation led the professor to declare, “AI-enabled cheating is steering society toward failure.”
The Genesis of the Incident
The decision by Serrano to use remote exams for his “ECON 1170” course was influenced by a tragic incident. In December 2025, a shooting occurred on Brown University’s campus, claiming two lives, including someone Serrano had recently met. Deeply affected, he opted to implement a remote exam format for both midterm and final exams in the spring semester of 2026.
Enrollment in the course surged as a result. Normally attracting fewer than 30 students, sometimes as few as eight, the class saw 86 students register. The midterm exam held on March 5 yielded unprecedented results: the average score hit 96 out of 100, with 40 students submitting flawless answers and the rest earning high marks.
Speaking to Inside Higher Ed, Serrano noted, “Historically, the average score for this course’s midterm exams ranged between 65 and 80. This time, I deliberately crafted a more challenging exam since remote testing offers unlimited time.” Yet, the scores far surpassed any previous records.
Suspicious Answer Patterns
The peculiarity extended beyond numbers to the content of the answers. Serrano observed an unusually “verbose and roundabout style” in the responses. When he and a graduate student inputted the same questions into ChatGPT, they found striking similarities between the AI-generated answers and those submitted by students.
Convinced of the likelihood of AI-enabled cheating, Serrano decided to revert to in-person exams for the final test. This marked a pivotal moment to evaluate the students’ genuine abilities without relying on AI. In an interview with El País, Serrano warned, “We cannot choose the path of becoming fools,” highlighting the societal risks posed by over-reliance on AI.
Dilemmas in the Education Sector
This issue is by no means unique to Brown University. A recent study at Princeton University revealed that 29.9% of students admitted to using AI for cheating on at least one exam or assignment. While Ivy League students are characterized by high intelligence, structural challenges such as time constraints and competitive pressures make them prone to shortcuts.
Advances in AI technology have made detecting cheating increasingly difficult. Professors are left to rely on subtle discrepancies in writing style and patterns to identify irregularities. Comprehensive measures, including institution-wide policies, are urgently needed. Although educators like Serrano who openly address such issues remain rare, this disclosure has sparked widespread discussion within the academic community.
Editorial Opinion
This case vividly illustrates the danger AI poses to the essence of “learning.” In the short term, universities may reconsider remote exams and accelerate the establishment of clear rules regarding AI usage. The adoption of AI detection tools in grading processes is also expected to gain momentum.
From a long-term perspective, dependence on AI could hinder students from developing critical thinking and problem-solving skills, potentially devaluing academic degrees altogether. Educational institutions must seek ways to integrate AI responsibly—fostering critical thinking rather than prohibiting its use outright.
As an editorial team, we pose this question: How should society balance the trade-off between the productivity gains offered by AI and the intellectual development of humanity? This revelation serves as a critical catalyst for such discussions.
References
- Ars Technica: “We cannot choose to become idiots”: The AI cheating scandal roiling Brown University — Published July 8, 2026
Frequently Asked Questions
- How did the professor at Brown University uncover AI cheating?
- The professor noticed an unprecedented average score of 96 in remote exams, with nearly half of the students achieving perfect marks. He also observed consistent anomalies in answer styles. When he tested the same questions using ChatGPT, the results closely resembled the students' submissions, prompting him to switch to in-person exams.
- Is AI-enabled cheating occurring at other universities?
- A study at Princeton University found that approximately 30% of students admitted to using AI for cheating. Similar trends are suspected across Ivy League institutions, with remote exams potentially serving as a breeding ground for misconduct. Universities are now under pressure to revise their policies.
- What measures can prevent AI-enabled cheating?
- Strategies include enforcing in-person exams, utilizing AI detection software, designing assignments that emphasize process-based grading, and incorporating oral examinations. Establishing guidelines for the responsible use of AI as a learning tool is equally vital. A dual approach combining technological solutions and educational reforms is necessary.
Comments