Why Banning AI Glasses Won’t Fix Exam Cheating: Lessons From South Korea for Parents and Teachers
The news alert appeared while I was waiting for coffee in the staff room. It was a gray Seoul morning, the kind when the heating smells like dust and everyone is already tired. A colleague held up her phone. “They’re saying students used AI glasses on the mock college entrance exam,” she said.
I leaned over, but I already understood the shape of it. Camera-equipped frames. A phone nearby. Questions photographed and sent to an AI chatbot. Answers returned through a hidden earbud.
In South Korea, the College Scholastic Ability Test does not feel like one ordinary school day. The country rearranges itself around it. Planes are grounded, offices open late, and younger students line up outside test centers handing out candy and good-luck notes. For many of my high school students, that pressure is not abstract. They return from hagwon lessons late, carry color-coded planners, and know their parents compare percentiles in group chats. So when a new tool appears, the question is not only “Could you get away with it?” It is also “How much would you risk to not lose?”
Two days later, I watched a student named Jiwoo adjust a pair of black glasses during a reading quiz. They looked ordinary, slightly smudged, the kind half the boys in the class wear. I almost said nothing. Then I noticed a tiny white reflection near the hinge. My chest tightened. I walked over and asked him to take them off.
He looked confused, then embarrassed. They were prescription glasses with a blue-light filter. The light was just the ceiling. I handed them back and spent the rest of the period feeling as though I had mistaken a kid for a suspect.
After class, a girl named Seoyeon stayed behind. “You know those glasses can’t really give you answers in here, right?” she said.
I asked what she meant. She looked at me like I was years behind. “You point the camera, it reads the question, and it talks through your earbud. But everyone knows the old ones have a light. The new ones don’t.”
She was not bragging. She was simply explaining the world, the way students do when they realize adults are still arguing over last year’s rulebook.
That conversation stayed with me. Our exam policy listed phones, smartwatches, cameras, and electronic communication devices. It did not clearly list glasses because, until recently, glasses were not a cheating problem. Next semester, the list could include rings, pens, buttons, or frames that look completely normal. We can ban each object as it appears, but we will always be one product launch behind.
This does not mean rules are useless. On closed-book test days, I still use a numbered phone pocket and a paper sign-in sheet. Earbuds go in bags. Camera-equipped frames are treated like phones, while ordinary prescription glasses stay on. The rule is simple, announced in advance, and applied to everyone. But I stopped treating detection as the main job.
Last semester, I changed my tenth-grade literature assessment. Instead of one in-class essay worth most of the grade, students did three shorter tasks. They annotated a poem in class for ten minutes and handed the paper to me. Then they wrote a draft on a school Chromebook using a simple template: claim, evidence, counterargument, final choice. During the next class, I held two-minute conferences while everyone worked. I would point to a line and ask, “Why did you choose this word?” or “What changed between your draft and final?”
If students could explain their choices, the grade stood. If they could not, we did a short reteach and revised the essay together. It took time, and some conferences were clumsy, but it made an AI-generated answer much less useful. A chatbot can produce a polished paragraph. It cannot sit across from a teacher and explain why a student picked that particular piece of evidence.
For homework, students were allowed to use AI to brainstorm or check grammar, but they had to paste the chat record at the bottom of the document. I did not grade the chat. I checked whether the final essay matched the in-class outline and whether the student could discuss it. That distinction mattered.
Parents can do something similar at home. During a conference, Minseo’s mother told me she checked his phone every night but still worried he would find another way to cheat. I suggested she stop asking “What did you score?” the moment he came through the door. Instead, she could say, “Show me one question you got wrong and tell me what you were thinking.”
That question does three small but useful things. It treats mistakes as information, it asks him to explain his reasoning, and it separates parental approval from the number. She tried it, awkwardly at first. Later she told me Minseo had talked about a geometry problem for ten minutes. He had not, however, wanted a hug afterward.
At home, exam morning works better when it is boring: breakfast, two pencils, a transparent pouch, and no last-minute predictions about what the score will mean. If a child wears smart glasses regularly, explain that on test days they are treated like a phone, not a style choice. If they need prescription glasses, do not make them feel ashamed for wearing them.
Schools also need to update acceptable-use policies so camera glasses and AI tools are named directly, but the rules should be short. A forty-page handbook will not be remembered. A poster by the door is more useful: phones in pouches, no smart glasses or earbuds, paper only, raise your hand to leave the room. Staff need training so device checks do not fall more heavily on certain students. When the rule changes, one sentence is enough: “This protects your work, not because we assume you cheated.”
The harder problem is the pressure underneath. A teenager who believes one score determines their future, their family’s pride, and their value will make choices in the moment that their adult self may regret. Policing can stop a pair of glasses. It cannot make a seventeen-year-old believe that one exam is not the whole story.
I see this even outside exam season. My niece, who attends a Seoul middle school, used to freeze before hagwon tests. Now she spends ten minutes making her own mini-quiz from questions she got wrong, then teaches me one grammar rule. If I can understand her explanation, she probably understands it too. No app is involved, and the rule is simple: mistakes become the next quiz.
Last week, Jiwoo came to class wearing the same smudged glasses. I did not look at the hinge. He put his phone in the pocket, sat down, and started annotating. Somewhere, a newer device is probably already on sale. I cannot police every object in a teenager’s life. But I can keep asking them to show me their thinking—on paper, in drafts, and out loud. It is not a perfect system. It is just slower than the glasses, and for now, it still tells me more about what they actually know.
Please indicate: Thinking In Educating » Why Banning AI Glasses Won’t Fix Exam Cheating: Lessons From South Korea for Parents and Teachers