In October last year, collective AI-assisted cheating occurred at several universities including Yonsei University (YU), Korea University (KU), and Seoul National University. These incidents highlight not only individual student misconduct, but also the limitations of university response systems and educational structures that have struggled to adapt to rapid technological change. The urgent need for the new education system based on AI usage as well as the clear guidelines for AI use has become increasingly evident.

​

ChatGPT on a phone screen 

Photo : Yonhap News (yna.co.kr)

The Reality of AI-assisted Cheating in Academia

AI-assisted cheating is rapidly emerging as a growing concern across universities.

On October 25, 2025, suspicions of collective cheating arose during the midterm exam of the online lecture “Interdisciplinary Understanding of an Aging Society” at KU. It was later revealed that some students had shared screenshots of exam questions through KakaoTalk open chat and used AI tools to search answers. In response, on November 27, KU announced through its Learning Management System (LMS) that the situation is too severe to overlook.

​

This was not the first such incident at KU. On October 20, a similar collective cheating was uncovered in a major course of the College of Engineering. During an online exam conducted through LMS, some students exploited a system loophole that let students repeatedly resubmit the answer after checking the answers. The professor in charge announced that some students had taken advantage of this flaw. To ensure fairness, the entire class was required to retake the exam in an offline, paper-based format.

​Soon after, a similar case occurred during the midterm exam of the course “Natural Language Processing and ChatGPT” at YU. Although prior preventive measures had been implemented, including requiring students to submit video showing their computer screens, hands, and faces simultaneously—were taken, some students repeatedly looked away toward blind spots out of the camera's view and used multiple overlapping programs on their screens. These actions were later revealed through video review.

​

Afterward, the professor stated that students who voluntarily confessed would receive a zero score for the midterm exam, while those who denied cheating despite clear evidence would face disciplinary action such as suspension.

​

Ongoing Controversies Over AI Usage, Yet No Clear Identification of Causes

The repeated emergence of AI-assisted cheating is not limited to a single university, but reflects a broader structural issue across whole university field. Many universities lack clear guidelines or regulations regarding AI use, leaving students confused about boundary between permissible assistance and cheating. A Comparative Analysis of Generative AI Guidelines in Domestic and International Universities published in August 2025 noted that domestic universities frequently emphasize ethical responsibility without offering sufficient practical action plan or educational materials.

​

Also the online exam's feature appears to reduce students' sense of tension and moral awareness, leading them to fail recognizing their actions as cheating. Kang Su-hyeon, student inh the Dept. of Mongolian at Hankuk University of Foreign Studies, recalled, "Taking exams at home was convenient, but it made me harder to perceive it as a real exam. It felt more like taking a practice exam, so I barely felt any tension." She added that "Being in a private space without any supervision may have influenced students to act immorally,” emphasizing how online exams can blur ethical judgment.

​

Some argue that AI-related cheating incidents reveal fundamental limitations of existing assessment systems. Professor Park Han-woo of Yeungnam University's Dept. of Media and Communication stated, "The controversy over AI-assisted cheating ultimately reveals structural flaws of universities' assessment systems. Current assessment system fails to reflect the reality that students' knowledge is already extended through AI and other digital tools."

​

Current Education System Faces Its Limits: Urgent Need for Structural Transformation

These incidents reveal that universities' existing response systems about AI usage have reached a structural limit. In this situation, the Ministry of Education and the Korean Council for University Education announced on November 27 that they had begun discussions on developing AI Ethics Guidelines for University Students. The guidelines are expected to include versions for both students and faculty, and are scheduled for release as early as March.

​

However, some warn that ethical guidelines may not be sufficient to fundamentally prevent AI-assisted cheating problem. Prof. Park noted, "Ethical guidelines can be helpful, but they are ultimately only guidelines. Practical solutions will come when evaluation systems integrate students' reasoning abilities with their use of AI tools," emphasizing the need for a structural transformation in assessment practices.

​

Meanwhile, needs for AI literacy education are also growing stronger. Prof. Lee Jae-yeon, director of the UNIST Center for Digital Humanities, emphasized, "Completely forbidding student's AI use is no longer realistic. We must teach students how to use AI properly." He added that "Since AI-generated information constitutes an intellectual copyright, it is essential to educate students how to mark source and citation when using AI derived content."

​

The AI-assisted cheating problem reveals not only the misuse of technology by students, but also the limitations of the current educational system that has failed to adapt to a rapidly changing society. Despite AI’s deep integration into everyday life, universities still rely on traditional responses such as surveillance reinforcement and outright bans. It suggests the need of fundamental transition toward an educational model that fosters creativity, responsible AI use, critical thinking, and ethical judgement.