The campus was thrown into shock after large-scale cheating surfaced in a class at Yonsei University. With many reportedly using Generative AI such as ChatGPT, debate over AI ethics on campuses is reigniting.
According to academia on the 9th, the professor in charge of the junior-level class Natural Language Processing (NLP) and ChatGPT at Yonsei University's Sinchon campus recently said that many cases of cheating were found and announced that the midterm exam scores of the students caught would all be marked zero. About 600 students are enrolled in the course, and the test was conducted remotely on the 15th of last month.
The exam was administered by logging into an online site and answering multiple-choice questions, and test-takers had to record and submit video showing their face and hands throughout the test. However, signs of cheating emerged as some students adjusted the camera angle or overlaid multiple programs on the screen. The professor was said to have posted a notice urging students to turn themselves in.
The exact number of cheating cases has not been disclosed, but among students, there was talk that more than half cheated. In a poll posted on the college community Everytime, 190 out of 353 respondents answered that they cheated.
Most students appear to have used AI during the test. Student A (25) said most used ChatGPT and noted that it seemed hard to get a good grade if only I didn't use it. A previous semester student, B, also said many friends searched with AI during the exam.
It has been three years since AI took deep root in academic settings, but universities' responses remain inadequate. According to a survey by the Korea Research Institute for Vocational Education and Training (KRIVET), 91.7% of 726 students at 4- to 6-year institutions said they had used AI for assignments or information searches, but a survey by the Korean Council for University Education found that 71.1% of 131 universities nationwide had yet to establish Generative AI guidelines.