As AI glasses get harder to detect, universities scramble to update exam rules

Visitors try on AI smart glasses jointly developed by Ray-Ban and Meta at KT Plaza in Gwanghwamun, Seoul, July 22. Yonhap

Artificial intelligence (AI) glasses are emerging as a new tool for academic dishonesty, with a string of recent cases exposing their potential for misuse, putting the education sector under growing pressure to respond.

The Ministry of Education recently instructed metropolitan and provincial education offices to classify AI glasses as prohibited items in examination rooms and closely monitor examinees for suspicious behavior, according to Yonhap News Agency, Sunday.

The directive follows a series of cases of cheating involving the devices. Earlier in May, three examinees wearing AI glasses were caught during the Industrial Engineer Electrical and Fire Protection Systems Engineer certification exams. Two test takers were later caught using the devices during an English language test.

Concern is mounting across the education sector ahead of the College Scholastic Ability (CSAT) Test and rolling admissions exams — high-stakes, state-administered tests that have drawn relatively close scrutiny due to their weight in Korea's highly competitive college admissions process. Even so, many institutions still lack clear measures to prevent AI-assisted cheating.

An even larger vulnerability lies within university-administered exams. Unlike the CSAT and other state-administered exams, academic misconduct in university exams is generally not subject to criminal penalties, leaving enforcement up to the institutions and prompting criticism that they have been slow to strengthen anti-cheating measures.

Korean universities already faced backlash last year after mass cheating cases exposed shortcomings in policies governing students’ use of generative AI.

Although many institutions later introduced AI-use guidelines, concerns persist that their countermeasures have failed to keep pace with the rapid evolution of the technology.

Seoul National University said it has yet to discuss how to respond to academic misconduct involving AI glasses. Korea University, Ewha Womans University and Sungkyunkwan University also reported that they have no dedicated guidelines in place, although they are reviewing possible countermeasures.

While some universities have begun introducing guidelines, their effectiveness is uncertain.

Chung-Ang University’s College of Medicine revised its student assessment guidelines in September last year to ban AI glasses from examination rooms. It also included instructions for exam supervisors to watch for suspicious behavior, including unusually thick eyeglass frames and examinees repeatedly handling their frames.

However, such measures may already be outdated as AI glasses continue to evolve rapidly.

Samsung’s latest AI glasses, unveiled on July 22, feature a display-free, voice based design with slim frames and a lightweight build, making them difficult to distinguish from ordinary eyeglasses.

The rapid advancement of AI devices is fueling concerns that increasingly sophisticated methods of cheating will continue to emerge.

AI-powered calculators equipped with generative AI and built-in cameras have already gone on sale overseas, while compact AI devices such as smart contact lenses are also under development.

Educators increasingly argue that the rapid evolution of AI requires a fundamental rethink of how students are assessed, rather than relying solely on tougher anti-cheating measures.