For grown-ups
Is AI safe for kids?
The honest answer is: it depends entirely on how it's used — and that's good news, because how it's used is the part you control. Here is what's genuinely fine, what deserves a watchful eye, and the few house rules that cover almost everything.
The short answer
AI is safe for young children the way a kitchen is safe: wonderful with a grown-up in the room, not a place to play alone.
Used together, for short and specific things — explain this word, tell us a silly story about a dinosaur, help us think of rainy-day games — it's genuinely useful and often delightful. The risks appear when it becomes a private, open-ended companion instead of a shared tool.
What's genuinely fine
Answering the hundredth “why?” of the day
Curiosity questions are AI's best use with kids. Why is the sky blue, how do bees sleep, what did dinosaurs eat — it never gets tired, and you can check the answers together.
Making up stories together
“A story about a brave squirrel who is afraid of heights” is a wonderful prompt to build with your child. They learn that they direct the machine — not the other way round.
Practising words and ideas
Explaining a new word, suggesting rhymes, inventing tongue twisters. Short, specific tasks where you can see exactly what came out.
What to watch
Three things deserve your attention, and none of them are reasons to panic — they're reasons to stay in the room.
It sounds sure even when it's wrong
A chatbot states a made-up fact in exactly the same friendly voice as a true one. Children believe confident voices — adults do too. The fix is a habit, not a warning: “let's check if that's true.”
It never says no, and never needs anything back
An always-agreeable companion is not how friendship works. If a child starts preferring the chatbot to people, that's the moment to step in — and to say plainly that a computer is never the one to go to when something is wrong.
It remembers what you type
Treat everything typed into a chatbot as if it were said in public. No full names, no school, no address, no photos, no family details — for you or your child.
Four house rules that cover almost everything
- 1. Together, in the open. AI time happens with a grown-up, on a screen everyone can see.
- 2. Short and specific. One question or one story, then done — not an open-ended afternoon with a chatbot.
- 3. Nothing personal goes in. No names, school, street, birthday, or photos.
- 4. Check one thing. Pick one answer and look it up together. It teaches the habit that protects them everywhere else too.
One more thing worth knowing: most chatbots' own rules say they're for ages 13 and up, some 18 and up. That's another reason the grown-up stays in the room — for young children, you are the app.
The goal isn't avoiding AI — it's a child who handles it well
Your child will grow up with AI the way you grew up with the internet. The children who do best won't be the ones who never touched it — they'll be the ones who learned early to ask where an answer came from, to check it, and to decide for themselves.
That's exactly the habit our stories practise. You can read more in our guide to explaining AI to kids.
Let a story do the first conversation
Our first book, Mila & Dot: Dot Draws a Very Strange Cat, puts these ideas into a story for ages 4–6 — Mila, her cat Pickles, and Dot, an AI helper who is helpful but very much not perfect. It ends with grown-up notes: a page of questions and activities so the conversation keeps going after the last page.
For little thinkers with big questions.