For grown-ups
How to explain AI to kids
You don't need to understand how a chatbot works to explain it to a five-year-old. You need one honest sentence, and three habits they can practise until they're their own. Here is exactly what to say, and what to keep an eye on.
Start with one sentence they can repeat
“AI is a computer that has looked at so many examples that it can guess what comes next.”
That's a real description, not a simplified one. A chatbot has read an enormous amount of writing and learned which words tend to follow which. It is not thinking, and it is not remembering the way your child does — it is matching patterns.
Then say the part nobody should skip: it can be wrong, and it sounds exactly as sure when it's wrong as when it's right. Kids who hear that early don't feel tricked later.
The three habits: Ask, Check, Choose
Facts about AI go out of date. Habits don't. These are the three we build every story around, and each one takes about a minute at the kitchen table.
1. Ask — where did this come from?
Before the answer matters, the source does. “Did a person write this, or did a computer make it up?” is a question a four-year-old can learn, and it never stops being useful.
2. Check — two minutes, together
Pick one thing the AI said and look it up side by side: a grown-up book, a museum website, a teacher, a park visit. You're not proving the chatbot bad — you're teaching that checking is just what you do after asking.
3. Choose — they decide
The AI suggests. Your child chooses. “Does that match what you already know? Do you want to use it, or try it your own way first?” A child who has practised saying no to an answer is a child who will say no to a bad recommendation at fourteen.
Is AI good for kids?
It's a tool with unusually good manners, so it feels like a friend and never gets tired of answering. That's the opportunity and the risk in the same sentence. (Wondering about safety specifically? Read our honest answer to is AI safe for kids?)
Used narrowly — explain this word, check this fact, help me practise — it's genuinely helpful and kids gain confidence. Used as an endless conversational companion or a homework factory, it quietly replaces the effort where the learning actually lives.
The practical guardrails
- • Keep it in the room you're in, on a screen everyone can see.
- • Most chatbots' own rules say 13 and up, some 18 and up — read the app's terms.
- • Nothing personal goes in: no full name, school, street, birthday, photos, or family details.
- • Short, specific sessions beat open-ended time with a chatbot.
What if they believe everything it says?
They will, because a confident voice is persuasive to an adult too. The fix isn't a warning about AI — it's a habit of checking that covers books, adverts, friends, and chatbots alike.
Make it a game rather than a lecture: “let's catch it being wrong.” Kids find mistakes fast, and once they've caught one themselves, the spell is broken for good.
Could it become too much of a friend?
Worth watching, especially in older children. A chatbot is always agreeable, never needs anything back, and never has a bad day — which is not how relationships work.
Say it plainly, once, and mean it: a computer is never the person to go to when something is wrong. If they're scared, sad, or in trouble, they tell a real grown-up — and a real grown-up will always thank them for it.
Let a story do the first conversation
Our first book, Mila & Dot: Dot Draws a Very Strange Cat, puts these habits 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.