How to Talk to Your Kid About AI

Start with how AI actually works

Before you talk about rules, give your child a working mental model. The version that lands for most kids: "AI is a really fast pattern-matcher. It read a huge pile of things people wrote on the internet, and now when you ask it a question it guesses the words that usually come next. That's why it's fast, why it sounds smart, why it's sometimes very wrong, and why it's not a person — even when it feels like one." Say it in your own words, and check for understanding by asking your child to explain it back to you. A kid who understands "guessing the next word" will not be surprised when the model makes up a historical event or invents a citation — and they'll be less likely to treat it as an oracle or a friend.

Talking points by age

Ages 5–8

Keep it concrete and use it together. AI is "a computer helper that can talk." Rules that work at this age: we use it with a grown-up, we don't tell it our real name or where we live, and if it says something weird or scary we tell a grown-up right away. Try it together on friendly, creative tasks — make up a bedtime story about the family cat, plan a pretend restaurant, invent silly animals. This is the age to build the habit of talking about AI, not to teach nuance.

Ages 9–12

Introduce the idea that AI can be confidently wrong. Try a fact-check game: ask a question you both know the answer to, then one you don't, and look up the second one together. Talk about privacy in specifics ("don't paste our address, your school name, or photos of yourself"), and about the difference between using AI to learn something and using it to skip learning it. This is also the age to name the emotional-support question out loud: "If you're upset, come to me or another adult first — the chatbot is not a person and can't actually help." Agree on where and when AI is okay (homework help at the kitchen table) and where it isn't (bedroom, late at night, alone).

Teens

Assume they're already using it — for homework, code, image edits, and increasingly for venting. Skip the lecture and get curious: which tools, for what, and what do they think the risks are? Cover four things explicitly: (1) academic integrity — what their school actually allows and how to cite AI use; (2) privacy — anything typed in can be stored and may be used to train models; (3) parasocial pull — chatbots are designed to feel warm and agreeable, which is why they can become the first place a lonely teen goes; (4) crisis — if they or a friend are struggling, the chatbot is not a therapist, and 988 (call or text, U.S.) or Crisis Text Line (text HOME to 741741) exist for a reason. Keep the conversation two-way — teens tune out one-way safety talks fast.

What to revisit over time

This isn't one conversation, it's a running one. Revisit it whenever your child changes schools, gets a new device, starts a new game or app, or when a story about AI is in the news. Quarterly is a good rhythm: five minutes at dinner about what they've been using AI for, what surprised them, and what felt off. Update your household rules as they get older — the deal you made at 9 should not be the deal at 14. Watch for the signs of unhealthy AI use, and let your child try AI in a safe setting like the Kid-Safe AI Playground.