AI Cheating in College: What a Professor’s Chart Reveals (And What Parents Need to Know)
The griddle was still smoking with leftover blueberry pancake batter last month when my sister Lila, a computer science professor at our local state university, pulled her laptop across my kitchen counter and tapped the screen. My 20-year-old son Jake had just left for his library study session, so we had 20 minutes of quiet before the brunch dishes piled up. The chart she pulled up was simple, just four sets of bar graphs comparing student AI use across three types of exams: proctored closed-book, in-person open-book, and take-home. It wasn’t fancy, just raw data from an anonymous survey she’d run that semester with 420 of her intro students. And it stopped me cold.
The tallest bar, for unapproved AI use on take-home exams, hit 61%. That’s almost two out of every three students. In-person open-book exams were next at 28%, and even proctored closed-book clocked in at 11%, mostly students sneaking phones out to check answers with chatbots. I leaned in, wiping syrup off my wrist. “You’re telling me most kids cheat with AI on take-home tests now?” I asked. She shook her head, tapping the second set of bars that broke down how students labeled their own behavior. Only 18% of the students who used AI without permission called it cheating. Most said it was “just extra help” or “the same as using Google to look up an answer.” That’s the part no one’s talking about, she said. The scale isn’t just about how many kids are doing it—it’s about how few of them see it as a problem.
I thought back to two weeks earlier, when I’d dropped off a late-night snack for Jake while he crammed for a biology midterm. His laptop had ChatGPT pulled up, and when I glanced over, he shrugged and said he was just checking a confusing cell process definition. I’d nodded and left, no second thoughts. Back then, I thought AI cheating was something that happened on big essays, not weekly midterms or problem sets. Lila’s chart blew that idea apart. She said she ran the survey after she noticed half her class got identical correct answers on a take-home problem set that she’d tweaked that year to be unsearchable on Google. When she asked around to other profs, most said they’d noticed the same thing, but no one had put hard numbers to it yet. The chart was just her way of showing the department how big the shift had been in just two years. Before 2023, less than 12% of her students reported unapproved AI use. Now it’s 61%.
We sat there for a minute, eating cold pancakes, talking about why this is happening. Lila said after COVID, most profs switched to more open-book and take-home exams to cut down on student stress and reduce cheating on proctored tests. No one anticipated that AI would make it so easy to get a full set of correct answers in 30 seconds. And kids who grew up with instant access to every answer online don’t see the line that older generations do. To them, AI isn’t a cheating secret—it’s just another tool, like a calculator or a textbook. The biggest gap, her chart showed, is between professor expectations and student understanding. At the bottom of the chart, she added a small extra bar: 82% of students said they’d never gotten a clear, explicit explanation from any of their professors that semester about what AI use is allowed and what isn’t. Most profs just leave the old “no unauthorized assistance” line in the syllabus and assume kids know what that means. They don’t.
After Lila left, I thought about what I could actually do about this, as a parent. Most of the takes I see online are either yelling at kids for being lazy or yelling at profs for being out of touch, neither of which helps me help Jake. I tried out Lila’s go-to opening question on Jake when he called that night to ask for grocery money. I didn’t lead with “are you cheating?”—that would have shut him down immediately. I just said, “Heard you have that stats midterm coming up. Did your prof clearly say what tools are allowed for it?” He paused for a second, then said, “Huh. I don’t think he ever actually said. The syllabus says ‘no unauthorized help,’ but everyone I know uses AI to check answers, so I assumed that’s fine.” There it was, exactly what Lila’s chart showed. He wasn’t trying to be dishonest. He just didn’t know the rules.
I told him to shoot his prof a one-sentence email: “Hi Professor, just wanted to confirm if using AI to check my work after I solve problems is allowed for the upcoming midterm. Thanks.” That’s it, no big speech, no awkward confrontation. He did it the next morning, and the prof wrote back 20 minutes later: AI isn’t allowed to generate answers, but it’s fine to use it to check your work after you try it yourself. Jake told me later that cleared up so much stress for him—he’d been worrying he was doing something wrong, but didn’t want to ask because he thought it would make him look like he was planning to cheat.
For teachers, Lila has made a small change that cut unapproved AI use in half in her own classes this semester: she lists all permitted and not permitted tools right at the top of every single exam and problem set, instead of just burying it in the syllabus. It’s one extra line, no extra work, and it eliminates all ambiguity. For parents, that one simple question—asking if the rules are clear—opens the conversation without putting kids on the defensive. It doesn’t assume your kid is doing something wrong, it just addresses the gap that the chart exposed.
Last week, Lila posted her chart online, and it got shared a few thousand times by education outlets. Some people called her alarmist, some people called all college students entitled cheaters. But the chart doesn’t judge, it just shows the scale of the shift we’re all in the middle of. No one has figured out the perfect fix yet. Jake still uses AI to study, now that he knows what’s allowed, and it’s helped him cut down on time he used to waste stuck on wrong problems. We don’t have this all figured out, and it’s still changing every semester as new tools pop up. But that small first step—asking about the rules, making the rules clear—has made more of a difference than any big lecture ever could.
Please indicate: Thinking In Educating » AI Cheating in College: What a Professor’s Chart Reveals (And What Parents Need to Know)