What I Learned About AI in Education As a Journalist and Parent of a Middle Schooler
Last October, I sat in the back of my 12-year-old son Leo’s pre-algebra class, notebook open, half-empty oat milk latte sweating through its paper cup. I was there reporting a story on how schools handle new AI tools, so I’d been warned half the kids would probably use it for their daily exit ticket. The prompt was a straightforward word problem: “If your family drives 320 miles on a road trip at 65 miles an hour, how long will it take you to get there, including a 45-minute lunch stop?” I leaned over a little, peeked at Leo’s Chromebook (the one with a cracked corner and a beat-up skateboard sticker on the back), and sure enough, he’d pasted the whole question into ChatGPT before I finished reading it myself.
After class, I cornered his teacher, Ms. Henderson, by the jar of lemon drops she keeps on her desk, expecting her to rant about cheating and wasted learning. Instead she popped a yellow drop into her mouth and shrugged. “I know they’re all doing it. I don’t care, as long as they can explain the answer to me when I call on them.” She’d changed her homework rules that fall: instead of 10 practice problems to turn in, kids only need to show full work for 3. If they used AI to get the answer first, they just have to write that down at the top of the page. That stopped me short, because up until that point, I’d spent a month reading doomsday op-eds about AI destroying education, and as a parent I’d been panicking right along with everyone else, sneaking checks of Leo’s homework folders to catch AI-generated text.
That night at home, I asked Leo why he used AI for a simple exit ticket problem. He was leaning against the kitchen counter eating a strawberry Pop-Tart, half-watching skate clips on his phone. “It’s just like when you use Google to find a quote for your article, Mom,” he said. “Why would I waste 20 minutes messing up long division when the robot can do it, so I can get to figuring out why the answer works that way?” It clicked. I use AI to crank out boring first drafts of meeting recaps and transcribe interview audio for my work all the time. I never thought of that as cheating; I thought of it as saving time for the parts of my job that actually need my input. Why was I holding Leo to a different standard?
Over the next two months of reporting, I talked to 12 teachers and 18 parents across three different school districts. Almost every parent I spoke to was panicking about cheating. Most had either caught their kid using AI or suspected they were, and almost all responded by banning phones during homework time, taking away devices, or making kids write all assignments by hand. What I found again and again was that this just made kids better at hiding it. One mom I met at Leo’s soccer practice told me her 10th grader got a zero on a history essay for using AI, so she took his Chromebook away for two weeks. A month later, she found he was writing all his essays on his old phone, clearing his search and browsing history every single night. He still hadn’t learned how to outline a thesis, he just learned how not to get caught.
The teachers and parents who weren’t pulling their hair out weren’t doing anything fancy or expensive. They just made small, concrete shifts that anyone can try at home or in the classroom. The first change they made was swapping one question for another: instead of asking “did you do this all by yourself?” they ask “what part did AI help you with, and what did you add that AI couldn’t do?” It’s a simple shift, it takes 10 seconds, and it turns an accusatory argument into a conversation about learning. The second shift was rethinking what homework is actually for. Homework isn’t meant to just generate correct answers; it’s meant to practice thinking. If AI can generate the correct answer, the new work becomes checking that answer, explaining the steps, or adding personal input that AI can’t. For example, when Leo had to write a paragraph about the climax of Wonder for ELA last month, I let him use AI to generate a basic plot summary to save time. Then he just had to add two sentences about how he’d feel if he was the new kid at school in that situation. It took me five extra minutes to check, and he actually engaged with the book instead of just copying a generic summary.
Last weekend, we split an $18 pizza between the three of us, and I asked Leo how much each person owed. He immediately reached for his phone, and I felt that old familiar twist of panic in my chest. Then I asked my new go-to question: “Okay, what’s the answer, and how do you know it’s right?” He paused for a second, thought it through, and said “Three times six is 18, so it’s six bucks each. AI said six too, so that checks out.” He knew how to do the math. He just didn’t want to, same as I use a calculator to add up my grocery receipt even though I know how to do addition by hand.
I still don’t have all the answers. No one does, really. We’re all figuring this out as we go. There are still things that worry me: what happens when he takes a standardized test where he can’t pull up AI? Will he have practiced the foundational skills enough when he gets to high school? I don’t know. But what I do know is that fighting AI, banning it, acting like it’s some big evil that rots kids’ brains isn’t working. Every kid I talked to for this story is using it anyway, they just get better at hiding it. Right now, the small, calm shift of asking what they did with the tool, instead of asking if they used it, is working for our family. It’s not perfect, and it’s still changing, but it’s a lot better than the constant arguments I was having six months ago.
Please indicate: Thinking In Educating » What I Learned About AI in Education As a Journalist and Parent of a Middle Schooler