Overview
- Korea University invalidated a large online midterm after students shared screenshots and answers in KakaoTalk open chats, with up to 500 participants in a course of about 1,400.
- The university is pursuing discipline, reviewing how to run the final exam, and has posted new AI usage guidelines from its Distance Learning Center.
- At Yonsei University, an online midterm for a 600‑student course on natural language processing was flagged for widespread misconduct using camera blind spots, multiple windows, and ChatGPT.
- About 40 Yonsei students have confessed to cheating, 10 more are suspected, and a student poll showed 211 of 387 respondents admitting they cheated.
- Yonsei will hold an emergency campus forum to address AI ethics and online assessment reforms, as surveys show heavy student reliance on generative AI and limited formal policies across universities.