AI Glasses in Exams: Why Policing Student Devices Isn’t Enough (and What Teachers and Parents Can Do)
Last Tuesday, I knelt beside Javi’s desk during a history quiz and pointed to the black frames perched on his nose.
“Are those regular glasses?” I asked.
He pushed them up with one finger. “Yeah, prescription. My eyes have been killing me.”
I nodded but stayed another second. They looked ordinary enough, except for a small pale light near the hinge. It might have been the overhead projector. Our school handbook says no phones, smartwatches, earbuds, or “any device capable of recording or communicating.” It does not really know what to do with glasses.
“If those connect to your phone or can take pictures, they need to go in your bag,” I said quietly.
He sighed, pulled a worn case out of his backpack, swapped to an older pair, and muttered something about how even his eyesight was suspicious now. I walked back to my desk feeling absurd. Ten minutes later, I watched a girl tuck hair behind her ear and wondered whether the tiny white bead in her ear was an earring or an earpiece.
That is the uncomfortable problem South Korea’s AI glasses cases pushed into public view. When I read the reports at my kitchen table that night, the details felt both extreme and familiar. Investigators said hundreds of people were accused of using camera-equipped glasses to stream exam questions to outside helpers, who sent answers back through tiny earpieces or linked devices. Some participants reportedly paid thousands of dollars for the scheme, which was tied to high-demand language and qualification tests.
My first thought was, That could never happen here. My second was, It already could, just on a smaller scale.
I teach two social studies classes and work with other teachers on assessment design. I am also the parent of a fourteen-year-old who once tried to hide a wireless earbud under his hood during homework because he “just wanted background music.” The line between helpful device, entertainment, medical tool, and cheating device has never been clean, and AI glasses make it nearly invisible.
A few years ago, exam day mostly meant checking for phones. Now we are checking watches, earrings, pens, calculator cases, hoodie strings, and glasses. Some of this is necessary. But treating every student like a possible concealment problem changes the room. Honest kids become tense. Teachers become suspicious. And young people who actually need assistive technology can be made to feel like rule-breakers.
The same pair of smart glasses that can help someone read signs, translate text, or record a lecture for later review can also stream a test page. A blanket rule against “all glasses” is not workable, and a rule against “anything suspicious” just invites bias and embarrassment. Schools have to be clearer about the behavior they are prohibiting: no recording, no outside communication, no live assistance during an assessment. Then they have to design assessments that do not depend entirely on catching the latest device.
After a particularly exhausting exam week in January, our department stopped trying to security-check every object and changed the test itself. Instead of fifty multiple-choice questions that rewarded memorization, students received two short historical sources, a graph, and a prompt asking which source was more reliable and why. They had to cite one line from each document. It was not thrilling, but it was hard to outsource to an outside helper staring at a livestream, because the answer depended on the papers in front of them.
We also made three versions of the assessment, but the versions were not just reordered questions. Names, dates, and source details changed. More importantly, every student sat with me or another adult for a ninety-second oral check sometime during the unit. I would pull a card and ask something simple: “Explain one cause of the price increase shown on that graph,” or “Tell me why the author of Document B might disagree with Document A.”
Those short conversations told me more than a metal detector ever could. A student might turn in a polished essay, but if he could not explain one sentence from it, I knew we needed a follow-up. I did not automatically accuse him of cheating. I asked, “Where did this idea come from?” Sometimes the answer was tutoring. Sometimes it was AI. Sometimes he understood it but froze when speaking. That distinction matters.
At home, I have tried to keep the rules equally plain. Phones, watches, and earbuds go in a basket near the kitchen door during forty-five-minute study blocks. My son uses a school-issued laptop for writing, and it stays on the dining table, not in his room. If he uses AI or a translator for something legitimate, he has to leave a note at the bottom: “Used translator for three vocabulary words” or “AI helped me make an outline, but I wrote the paragraphs.”
That last rule came from a science project where I asked him to explain a sentence he clearly did not write. He shrugged and said, “The website said it better.” I told him AI could help him understand a confusing concept, but it could not be the voice he submitted. Then I had him close the laptop and teach the concept to me in two sentences. He stumbled, tried again, and got it roughly right. That stumble is the part we are trying to protect, not punish.
Before a quiz, I no longer ask, “Are you ready?” The answer is almost always yes or a grunt. I ask, “What’s the one thing you hope isn’t on this test?” Then we spend ten minutes working on that exact thing. It gives him a way to admit uncertainty without feeling like he has already failed. I have also told him directly: “If you ever feel so trapped before a test that you think about hiding a device, come talk to me before the test. A bad score is fixable. Lying about what you know makes it harder for me to help you.”
None of this removes the pressure. Teenagers can understand consequences in the abstract and still choose a shortcut when grades, parental approval, social status, and fear all show up at once. The South Korea cases were not only a technology story. They were a high-stakes testing story. When one exam can decide a university place, a scholarship, or a job credential, people will use whatever tool they think gives them an edge. Banning the tool after it appears does not remove that pressure.
Schools can still set reasonable boundaries. Our classroom basket holds phones and smart watches. Students who need medical or assistive devices work with the nurse and counseling office so accommodations are arranged privately. During assessments, smart glasses stay in bags unless a documented accommodation says otherwise. We do not inspect ordinary frames or make students prove their prescriptions. We focus on what students do, not whether their glasses look unusual.
But the deeper fix has to be assessment. If a test can be answered by photographing a page and waiting for a reply, the test is too easy to copy. If a homework grade depends mainly on a final document with no visible thinking, AI will do too much of the work. In-class writing, drafts, source comparison, oral explanation, and questions tied to specific classroom material give teachers evidence of learning that no outside helper can easily manufacture.
Last week, Javi walked into class, pointed to his glasses before I could say anything, and said, “Old pair. No camera. No Wi-Fi. Just terrible vision.”
“Thank you,” I said.
It was a small moment, and it did not solve anything. Some students still forget the basket. Some parents ask why devices cannot simply be banned completely. Other teachers worry about students who find tools we have not heard of yet. I do not have a clean answer to that. But I do know I would rather build a test that shows me how a student thinks than spend the year searching glasses, earrings, and pencil cases.
The technology will keep changing. The exams have to change too.
Please indicate: Thinking In Educating » AI Glasses in Exams: Why Policing Student Devices Isn’t Enough (and What Teachers and Parents Can Do)