What Happens When Kids Use AI? A Parent’s View From the Kitchen Table
The first time I saw my daughter’s screen light up with a response from ChatGPT, it was mid-afternoon on a Wednesday. She was nine, sitting at the kitchen table with her math worksheet, and instead of writing in the boxes, she had typed the problem into the chat window. What is 3/4 + 1/6? Please show the steps.
I remember the flash of irritation. “You have to do that yourself,” I said, reaching for her tablet.
She looked up at me, not guilty in the least. “I just want to see how it works.”
That was about eight months ago. Since then, I’ve learned that “I just want to see how it works” covers a lot of territory. She uses AI for jokes, for story ideas, to explain why volcanoes erupt, and occasionally to check her math homework. Sometimes I catch her asking it questions about her friends—like what to do when someone doesn’t invite you to a birthday party. And last week, I found she had asked it to write a paragraph comparing two characters from a book her class was reading. A paragraph she then copied into her reading journal, word for word.
That last one was the moment I stopped being curious and started being concerned.
I didn’t yell. I didn’t take away the tablet. I sat down across from her, picked up the notebook, and read her the paragraph out loud. Then I asked, “What do you think about this?”
She shrugged. “It’s better than what I would have written.”
“Maybe,” I said. “But it’s not yours. And if I’m reading it in your notebook, I’m reading something you didn’t actually write.”
We sat with that for a minute. I could see her processing it. She wasn’t being defiant. She genuinely thought that if the answer was right and the writing was good, that was what mattered. The idea that the work itself—her own thinking and words—had value beyond the final result hadn’t occurred to her.
That conversation, and the ones that followed, taught me more about AI and kids than any parenting article I’ve read since.
The first thing I had to accept was that banning it wasn’t realistic. She’s ten. By middle school, it will be woven into everything. Her school is already talking about AI policies for next year, which means the teachers are in that same messy middle. They don’t know what it looks like when a kid learns on their own terms anymore. Neither do I.
So instead of making rules from a place of fear, I made them from a place of curiosity. I started asking her to show me what she was doing. Not to catch her, but to understand.
Here’s what I learned by watching her.
Kids don’t use AI the way adults assume they do. My daughter mostly treats it like a search engine that talks back. She asks it questions, gets answers, asks follow-up questions. Half the time she’s not trying to cheat—she’s trying to learn something on her own, at her own speed, without a parent or teacher watching her. That’s not lazy. That’s agency. It’s also something we should probably pay attention to, because it’s the first time a generation has had an interactive tutor that never gets tired or annoyed.
But the other half of the time, she’s looking for the shortest path to done. Because she’s ten, and ten-year-olds are wired for efficiency. Who can blame her? If a tool can do something for you, you use it. That’s what tools are for.
The key, I realized, was teaching her to distinguish between having the work done and doing the thinking. That’s a much more subtle distinction than “AI is bad” or “AI is good.” It’s the distinction between using a calculator to check that you did the math right, and asking the calculator to give you the answer without doing the math at all. Both involve a machine. Only one involves learning.
We settled on a few simple rules that have actually held.
First draft is hers. Whatever she’s writing for school—a paragraph, a story, a science response—she has to draft it in her own words first. Then she can use AI to check it, improve it, or ask follow-up questions. But the foundation has to be hers. This means her writing will sometimes be messier, rougher, less polished. That’s fine. That’s real.
Show me what you did. Whenever she uses AI for schoolwork, she has to show me both her version and the AI version. We talk about what the AI did differently and whether that difference is actually better or just cleaner. Half the time, it’s just cleaner. Her own voice gets lost in the improvement. I point that out to her. I want her to recognize that “sounds smarter” isn’t always “thinks better.”
Ask it to explain, not just answer. This one came from watching her work. When she asks AI to solve a math problem, she now has to ask it to explain the steps in her own words back to her. If she can’t explain it, she doesn’t understand it. I don’t care if it’s “correct.”
We also started doing what I call shared sessions. Once a week, we sit with the tablet open and ask AI things together. Not for homework—just for fun. We’ve planned a fake vacation itinerary, asked it to invent a new flavor of ice cream, and had it explain why dogs tilt their heads. This does two things. It keeps AI from feeling like a secret, and it shows her that I’m not worried about it. I’m learning with her.
The teacher response helped too. I emailed her teacher and asked, honestly, what she thought about kids using AI at home. Her response was surprisingly relaxed. She said it was inevitable and that what she cares about is whether a child can talk through their own work. She’d rather see a messy paragraph and hear the kid explain her reasoning than see a clean summary with no understanding behind it. That gave me a clearer sense of where the boundary should be: AI can help with the product, as long as the process belongs to the child.
None of this is settled. I still check her tablet periodically. I still find things that make me pause. Last week she asked AI, “Why does my mom get annoying when she checks my work?” I didn’t read further. I closed the screen and took a breath. I figured if she’s asking AI for parenting advice instead of asking me, at least the chat logs are saving me from saying something I’d regret. Maybe that’s the real lesson—they’re going to use the tools they have. Our job isn’t to pretend we can stop them, or that the tools are entirely good or bad. It’s to be curious enough about what they’re actually doing to stay in the conversation.
It’s not tidy. It’s messy. But it’s real, and it’s happening at kitchen tables everywhere. The best I can do is keep sitting at ours.
Please indicate: Thinking In Educating » What Happens When Kids Use AI