Recently, Lee Jin-woo, 32, a perfect scorer on South Korea’s Legal Education Eligibility Test (LEET), the law school admissions exam, was found in cardiac arrest at a motel in Seoul and confirmed dead. He had been mired in suspicions of AI cheating. Before his death, the perfect scorer once hailed as an “academic legend” had publicly said he would start a YouTube livestream on the agreed date in the morning and answer the Korean-language section of South Korea’s College Scholastic Ability Test (CSAT) live, using real-time answers to prove his innocence. Many netizens waited in the livestream room for verification. But when the stream was due to begin, he never appeared and his phone was turned off. The self-proof never took place.
The controversy began with a series of AI exam-cheating scandals exposed in South Korea. Netizens therefore suspected his perfect score was “irregular,” and accusations kept spreading online, bringing enormous public pressure on him. Police later confirmed his death; a suicide note was found at the scene, saying, “My body and mind are utterly exhausted.” The note neither admitted cheating nor offered an apology. With his death, the criminal case over suspected AI cheating no longer met the conditions for prosecution under South Korean law, and the investigation was terminated.
Beyond this individual case, the larger context is that AI is shaking South Korea’s exam system: similar cases have occurred frequently recently, and the top scorer in the tax accountant exam has publicly admitted to cheating with AI. South Korea’s education authorities were forced to urgently update exam-room bans, listing smart devices such as AI glasses as prohibited items. A ban on paper can stop devices, but it cannot stop technology from reshaping exam fairness. How to verify “genuine ability” in an exam environment with no proctoring blind spots has become a harder question to solve than assigning blame.




