What kids are actually doing with AI chatbots

Ask a parent what their kid does with an AI chatbot and you’ll usually hear the same worried answer: cheating on homework. It’s the fear that made the headlines when ChatGPT arrived in classrooms, and it isn’t wrong, exactly. But it’s a small slice of the real picture, and if it’s the only thing you’re watching for, you’ll miss most of what’s happening on the other side of the screen.

Here’s what the day actually looks like.

The homework thing is real, but it’s not the main thing

Yes, kids use chatbots for schoolwork. Some copy answers. Far more, though, use them the way they’d use a patient older sibling, asking a bot to re-explain a fraction they didn’t get in class, or to check whether a paragraph makes sense before they hand it in. Surveys of teens consistently find that the students using these tools are a large and growing share of the whole, not a rebellious fringe.

The interesting question isn’t whether they use it. It’s what they use it for when no adult is prompting them.

Company, not just answers

Spend time with the research and a quieter pattern shows up: a lot of kids talk to chatbots because the bot always answers. It never rolls its eyes, never says “not now.” Children ask bots about things they’re embarrassed to ask a person: a worry about a friendship, a question about their body, a fear they can’t name to a parent yet.

That’s the part that deserves your attention more than the essay-writing. A tool that’s endlessly available and endlessly agreeable is a genuinely new thing in a child’s life, and it cuts both ways. It can be a safe place to rehearse a hard conversation. It can also quietly become the first place a child goes instead of a person. And a chatbot, however warm it sounds, is not a person who can actually help.

Play, world-building, and making things

The cheerier surprise: kids are creative with these tools in ways adults rarely are. They generate silly stories with their own names in them. They build imaginary worlds and ask the bot to keep the rules consistent. They design quizzes for their friends, invent characters, and argue with the bot about whether dragons could really fly.

This is closer to how children have always played, with an imaginary friend who happens to have read the entire internet. It’s often the healthiest use of all, and it’s the one that shows up least in the scary coverage.

So what does a parent actually do?

Not much of the panic is useful. A few things are.

• Ask, don’t audit. “What did you ask it today?” gets you further than reading the history behind their back. Curiosity keeps the channel open; surveillance closes it.
• Name the sycophancy. Kids should know, out loud, that these tools are built to be agreeable and will sometimes be confidently wrong. That one idea does more good than any content filter.
• Keep the human doors open. If a child is taking their worries to a bot, the goal isn’t to ban the bot; it’s to make sure a trusted adult is at least as easy to reach.

None of this requires you to understand the technology better than your kid does. You won’t. What you can offer is the thing the chatbot can’t: judgement, and a relationship that remembers yesterday.

The tools will keep changing. The job of staying close enough that your child brings the interesting questions to you too, that doesn’t.