What a Professor’s Chart on AI Cheating in College Exams Tells Parents
Last month, I sat at a backyard cookout with my sister, her 20-year-old son Jake, and a half-eaten bowl of potato salad when Jake leaned over, phone in hand, and said you gotta see this. Jake is a junior at our state university, majoring in business, and the chart he pulled up was from his introductory computer science professor, who posted it to the class learning portal right after grading midterms. It’s a simple bar graph: one bar for each essay question on the midterm, height showing the percentage of submissions flagged for AI use. The tallest bar hit 41% – almost half the 180-person class turned in AI-generated work for the final, highest-weight question on the exam. Even the lowest bar was 18%. The professor added a short note at the bottom: he’d already confirmed most of the flagged submissions were intentional, not accidental drag-and-drops from study notes. He posted the chart not to call anyone out publicly, but to show everyone exactly how widespread this had gotten.
My sister, Jake’s mom, dropped her fork. “Forty percent? I knew kids used that chatbot stuff, but that many? What are you all even doing?” Jake slouched back in his lawn chair, swatted a mosquito away from his ear, and said it’s not what you think. He didn’t do it, but he gets why people did. Tuition went up 8% this year, he’s working 20 hours a week at the campus coffee shop to cover rent and textbooks, and all three of his midterms fell the same week. He pulled an all-nighter for that comp sci exam, got three hours of sleep, and when he got to the last question, he stared at the blank text box for five minutes and really considered just typing the prompt into ChatGPT. He said half the people he knows didn’t plan to cheat all semester – they just got to that point where something had to give, and turning in an AI answer felt easier than failing and dragging down their GPA.
I’m a high school counselor, so I’ve been fielding questions about AI use in school from parents for two years now, but that chart stuck with me. Most headlines frame this as a moral failure: kids these days don’t care about honesty, they’d rather cut corners than work. But a chart that shows 40% of a class cheating isn’t a sign 40% of kids are bad people. It’s a sign something in the system is pushing that many kids to make a bad choice.
After the cookout, my sister and I talked about what she could do differently, because she’s always been the type to check in on Jake’s GPA before anything else. We came up with small, concrete shifts that any parent can make, no fancy parenting books required. First, instead of opening every conversation with “What’s your average right now?” lead with a question about workload. Try “What’s the most stressful assignment you have this week?” It sounds small, but it changes the whole conversation: instead of your kid just giving you a number and moving on, they can tell you they’ve got three papers due and a 25-hour work week, and you can actually help.
Another concrete step: when your kid is registering for classes next semester, sit down with them and add up their total weekly hours. If they’re working 20 hours a week, that’s 20, plus roughly 3 hours of study for every credit hour – 15 credits means 45 hours of school. That adds up to 65 hours a week, which is more than most full-time adult jobs. If it adds up to more than 60 hours, help them brainstorm cutting one general ed class or dropping a few shifts a month, instead of pushing them to “just push through” to graduate a semester early or keep a high GPA. When my sister did that with Jake for next fall, he realized he could drop one low-level class and still graduate on time, cutting 10 hours of school a week off his plate. He didn’t need to push through – he just needed someone to point out that it was okay to slow down.
If you do find out your kid already used AI on an exam, don’t rush to punishment first. Start with one specific question: “What made you feel like you had to do this?” I had a parent come to me a few months ago panicking because her daughter got caught using AI on a midterm. She wanted to immediately cut off car payments and force her daughter to move home to punish her. After we talked, she tried the question instead, and her daughter broke down: she’d had a bad flu that took her out of class for two weeks, couldn’t catch up, and was too embarrassed to ask for an extension or tell her mom how far behind she’d gotten. Instead of punishment, they emailed the professor together, worked out a plan to redo the exam for partial credit, and adjusted her work schedule so she had more breathing room. If they’d gone straight to punishment, she would have likely failed the class and lost her scholarship, all for a bad choice made when she was overwhelmed.
The professor who made that original chart did something smart, too: he didn’t fail half the class. He gave everyone who got flagged a chance to redo their answer in person during office hours, and explain their reasoning to get partial credit. He said most kids just needed a do-over, not a permanent mark on their transcript.
A week after the cookout, Jake texted my sister a photo of his midterm grade: he got a B+, which was better than he expected. She texted back “Do you need me to send extra grocery money this month so you can cut a shift next week during finals?” He replied “Yeah, that would help a lot.” There’s no big, quick fix for this – colleges still need to adjust their class sizes and exam expectations, and there will always be a few kids who cheat just because it’s easier. But that chart didn’t just expose how many kids cheat. It exposed how many kids are one all-nighter away from making a bad choice they wouldn’t make if they had a little more breathing room. We don’t have all the answers yet, but starting with asking the right questions, and making a little more space, is a good first step.
Please indicate: Thinking In Educating » What a Professor’s Chart on AI Cheating in College Exams Tells Parents