Why South Korea’s AI Glasses Exam Cheating Cases Show Banning Devices Isn’t Enough
A tenth grader named Min pushed his lunch tray back and leaned across the table. “You see the Korean exam thing? Glasses with AI.”
Rain was streaking the cafeteria windows, and I was carrying a stack of vocabulary quizzes that still smelled like copy machine ink. I stopped. “What glasses?”
Three kids talked over one another. A camera in the frames. Pictures of the test questions. Someone outside receiving them and sending answers back through a hidden earpiece.
“They made everyone turn in phones and smartwatches,” Min said. “Nobody said glasses.”
A girl across from him looked up. “That’s not even clever. That’s just expensive.”
By the last bell, the story had moved through the teachers’ room too. Mr. H was revising the exam-day announcement to include “smart glasses.” Someone asked how we were supposed to tell camera glasses from ordinary prescription glasses. Another teacher pointed out that students already wore hoods with earbuds tucked in, smartwatches under sleeves, and phones in socks. Last month, a boy had been sent to the office because his calculator seemed to have a note function. It did not. He was just nervous and kept pressing buttons.
That is the uncomfortable part of the South Korean cases: the problem was never only the glasses.
According to reports, South Korean investigators looked into several students accused of using AI-equipped glasses during the national college entrance exam, the CSAT. The glasses allegedly let them photograph test questions and relay them to people outside, including a private tutoring academy, while answers were sent back through concealed audio devices. The exam has long had strict rules against phones, smartwatches, tablets, and other electronics. But those rules named specific objects. They did not clearly anticipate ordinary-looking eyewear that could record, communicate, and tap into AI.
A few days later, education officials added smart glasses to the banned list. That is the usual pattern: a new device appears, a rule is updated, a new kind of inspection begins. But the list only grows longer. Phones. Watches. Earbuds. Glasses. Soon it may be rings, pins, pens, or something designed to look like a medical device.
I have sat through enough exam-day meetings to know how this goes. We buy phone pouches, post signs, remind students to empty their pockets, and then scan the room for anything unusual. Some of that is necessary. A standardized test does require basic security. But when a school’s main answer to cheating is “find the device,” it creates a strange atmosphere. Teachers become part detective. Students become suspects. Ordinary gestures—adjusting glasses, rubbing an ear, looking at the clock—start to feel suspicious.
That does not mean children should be left alone to use whatever tool they want. It means policing gadgets alone will never get ahead of the problem.
At home that night, I mentioned the story to my son while he picked at leftover pasta. He is fourteen and treats every new cheating story like a product review.
“Couldn’t someone just use a second laptop?” he said.
“During a test?”
“Not during class. If the test questions are online or everyone knows the worksheet bank.” He shrugged. “The rule says no phone. It doesn’t say no old laptop in a backpack.”
He was not bragging. He was describing the way many children now see school rules: not as fixed moral lines, but as systems to be worked around. The technology changes faster than the handbook, and students know it.
The deeper issue is pressure. In South Korea, the CSAT can shape university admissions, career paths, and family expectations. The country’s private tutoring industry is enormous, and many students study late into the night. When a single score carries that much weight, an expensive cheating device can feel like a rational risk to a frightened teenager. It is not a surprise that some people saw an opening.
Schools elsewhere have their own versions of that pressure: AP exams, honors placements, SAT and ACT scores, gifted-program tests, and the constant hum of college anxiety. Parents may say they value effort, but children hear what gets celebrated at dinner. If the only result that earns excitement is a high score, the message is clear enough.
So what actually helps?
In my classroom, I have stopped relying on one sealed, high-stakes test as the only evidence of learning. I still give short closed-note quizzes, but part of the assessment is open-note and asks students to explain their reasoning. A multiple-choice answer alone is worth very little if they cannot show why it fits. For writing, I collect an early handwritten brainstorm, a draft, and a final version. During work time, I pull aside three or four students at random and say, “Show me where this idea came from” or “Explain this paragraph back to me.”
That does not make cheating impossible. Nothing does. But it makes a hidden answer feed less useful. If students have to produce steps, corrections, and explanations, a stranger sending instant answers cannot carry the whole task.
At home, parents can lower the temperature in small, concrete ways. Before a big test, try saying: “After it’s over, I want you to teach me one question you got right and one you got wrong. I don’t need a perfect score.” During study time, have a basket by the front door for phones, watches, and earbuds. If a child needs a laptop, keep it in a common space and use the school’s allowed tools only. Do not interrogate their glasses; that is not the point.
When a cheating story appears in the news, use it as a conversation rather than a lecture. Ask, “What do kids at your school say about AI and tutors?” “Would someone tell a teacher if a friend was doing this?” “What would make cheating feel like the only choice?” Then listen. The answer may be uncomfortable. It may be “college,” “my parents,” or “everyone else is doing it.” Those answers tell you more than another banned-device list ever will.
Schools also need clearer rules about what AI can and cannot do on each assignment. “Do your own work” is no longer specific enough. A teacher might say, “For this essay, you may use AI to brainstorm, but not to write the body paragraphs. Submit your first prompt and response.” Another assignment might be closed-AI, just as some tests are closed-book. Students do not only need threats. They need to know exactly which tools are allowed for which kind of thinking.
None of this removes the need for consequences. Real cheating should be addressed, and students who break exam rules should face them. But a zero or a suspension does not answer the question that drove the choice: why was the score worth more than the risk?
This week, my advisory helped make a new exam-day checklist: phones in pouches, watches in bags, earbuds out of the room, regular glasses fine. A student raised his hand.
“What if the glasses look normal but have AI in them?”
I said that was exactly why we were changing some of our assessments too. He did not look fully reassured. Neither was I.
The pouches will go up next month. I will keep doing quick oral checks, collecting drafts, and giving credit for corrected mistakes. Somewhere, though, a new device is already being unboxed. The real question is whether schools keep buying better locks, or finally stop treating the lock as the whole solution.
Please indicate: Thinking In Educating » Why South Korea’s AI Glasses Exam Cheating Cases Show Banning Devices Isn’t Enough