The advent of AI in education is reshaping the concept of cheating, causing disruptions in American universities. AI technology is increasingly embedded in learning, leading to more extreme and inconsistent classroom practices aimed at curbing dishonesty. False accusations of cheating have risen, and the definition of academic cheating is evolving, according to professors, students, and academic integrity experts.
Monitoring Practices in Exams
At UCLA, some students reported receiving instructions to use a mirror large enough to show their entire workspace during online exams, while their laptop cameras were activated for supervision. In other cases, students were required to keep their arms crossed or behind their heads during video exams to prevent them from engaging with AI platforms. A Los Angeles attorney handling college disciplinary cases revealed that about 35% of her cases now involve AI-related accusations, with this figure rapidly growing. This situation is further exacerbated by inconsistent definitions of cheating and varied rules on AI usage.
Shifts in Cheating Practices
Uncertainty and contradictions characterize AI policies on campuses, with individual instructors often left to manage the ethical use of AI while maintaining academic integrity. For instance, some faculty allow AI to be cited in drafts, whereas UC Berkeley’s law school has implemented a near-total ban on AI usage. Lee Rainie, director of Elon University’s Imagining the Digital Future Center, stressed that trust issues are central to the conflict, as mutual suspicions arise between students and faculty.
Statistics on AI Usage
Igor Chirikov, a researcher at UC Berkeley, identified a significant increase in AI usage among students. A study he co-led surveyed over 95,000 students across 20 universities, revealing that two-thirds used AI for classwork, with one-third doing so regularly. A portion of these students employed AI in ways considered inappropriate, evidencing a shift in cheating dynamics. However, the challenge lies in proving AI-related dishonesty.
Issues with AI Detection Tools
Detection tools like Turnitin have claimed a low error rate, yet false positives remain a risk. Various studies report higher failure rates, particularly among non-native English speakers. UCLA student Titi Olotu, with specific accommodations, faced difficulties due to stringent AI monitoring, ultimately dropping a course due to its proctoring methods.
Student Experiences
Students have developed strategies to avoid AI accusations. Ivan Ornelas, a recent UC San Diego graduate, used Google Docs to maintain a clear version history and removed AI-like elements from his writing. Aldan Creo, another UC San Diego student, altered his explanations on assignments to appear less polished to prevent AI suspicions.
Legal and Institutional Responses
Attorney Adrienne Hahn has noted rapid growth in cases involving AI accusations, advising students to keep evidence of their work. Despite these challenges, very few students express outrage at accusations, according to Tricia Bertram Gallant from UC San Diego, who blames the lack of cohesive policy adaptation for much of the current turmoil.
Future Directions
Faculty have found diverse ways to adapt. Adam Kaiserman of College of the Canyons offers extra credit for students who avoid using phones during class. He has also adjusted his assignments to avoid tasks that AI could easily handle. Gallant reinforces that the core issue is not the prevalence of false accusations but rather the struggle of academic systems to keep pace with technological advancements.