How to Talk to Kids About AI: An Age-by-Age Guide With Scripts

Your child just asked you what AI is. You want to give them a real answer: not a dismissal, not a lecture, and definitely not something that scares them. Here is how.

To talk to kids about AI, match the explanation to their age. With 4- and 5-year-olds, describe AI as a computer program people made that can talk but is not alive. With 6- to 8-year-olds, explain that it learns from lots of examples and can get things wrong. With 9- to 12-year-olds, add bias, privacy and honesty about homework. With teenagers, talk about how language models work, deepfakes, and their own choices. Keep it short, honest and ongoing. Below are scripts for each stage, the questions children actually ask, and activities that make the ideas stick.

If you would like a refresher on the basics before you start, our plain-English parent's guide to understanding AI covers what AI is and how chatbots work.

Why This Conversation Matters

AI is already part of your child's world. It recommends what they watch, powers voice assistants, appears in apps and games, and increasingly shows up at school.

Children who understand what AI is, even at a basic level, make better decisions about using it. They are less likely to over-trust it, less likely to be frightened by it, and more likely to treat it as a tool rather than as magic or a person. UNESCO's AI competency framework for students sets out what children should learn about AI, from how it works to using it responsibly.

The conversation does not need to be technical. It needs to be honest, age-appropriate and ongoing.

Ages 4-5: AI Is a Computer Program That Can Talk

Children this age think in concrete terms. Connect AI to things they already know.

How to explain it

"You know how we can ask the speaker to play a song, and it does? People built a computer program that can listen to our words and answer. It isn't alive. It's a machine that's very good at following instructions."

"Some apps can answer your questions out loud. They're made by people to help children learn. But they aren't people, and they don't have feelings."

What to emphasise

• AI is a tool, like a calculator or a book. It helps us, but it is not alive.
• People built it and decided what it should do.
• It can get things wrong, so we check with a grown-up.

Questions they ask, and simple answers

"Is it real?" "It's real like a toy is real. Someone made it. But it doesn't think or feel like you and me."

"Can it hear me?" "When we press the button and talk, yes. That's why we only use apps made for kids, and why we never tell it where we live."

"Is it smart?" "It knows how to put lots of words together. But it can't feel happy or sad, and it doesn't know you the way your friends do."

At this age, any AI use should be short, with you beside them.

Ages 6-8: AI Learns From Examples

Children this age understand cause and effect, and many are starting to use apps with some independence.

How to explain it

Use the learning analogy. Children know what learning feels like.

"Imagine showing a computer a million pictures of cats and a million pictures of dogs, and telling it which is which each time. After a while it gets good at spotting the difference, even in a new picture. That's how AI learns: not by thinking, but by seeing lots and lots of examples."

"When you ask an AI a question, it puts together an answer from patterns in everything it learned. Most of the time that works well. Sometimes it gets it wrong, and it can sound just as sure when it's wrong."

What to emphasise

• AI learns from examples (data), not from experience
• It can be wrong, so important things get checked
• It does not understand feelings the way people do
• People make choices about what it learns

Conversation starters

• "What do you think would happen if we taught an AI with only wrong information?"
• "What's something a person can do that a computer never could?"
• "If you could teach an AI one thing, what would it be?"

Try this together

Let your child ask an AI a question they already know the answer to. Did it get it right? Was anything missing? This builds the habit of checking, early.

Ages 9-12: AI Has Limits and Biases

This group can handle nuance. They are likely to meet AI at school and through friends, and they are forming habits that will last.

How to explain it

"AI works by finding patterns in huge amounts of text and pictures. If the examples it learned from were unfair or wrong, its answers can be unfair or wrong too. It doesn't know what's true. It knows what's likely to come next, based on what it saw."

What to emphasise

• AI reflects the data it was trained on, including biases and errors
• It generates likely answers, not guaranteed facts
• Copying AI homework is not learning; using AI to understand can be
• What you type or say to an AI can be stored

Topics to discuss

Deepfakes. Children this age are on video platforms. They need to know AI can make realistic fake images, voices and videos. Look at some examples together and talk about how to spot them.

Honesty at school. Be direct: having AI write your work is not learning, and it is not honest. Using it to understand a tricky idea, then doing the work yourself, is different.

Privacy. Explain that conversations with many AI tools are stored and may be used to improve them, which is why names, addresses, school details and photos stay private.

Try this together

Have your child ask an AI about a topic from school, then check the answer against a textbook or trusted website. What did it get right, what did it miss, and what did it get wrong? Our guide to teaching kids critical thinking with AI has five more exercises like this.

Ages 13+: Ethics, Opportunity and Responsibility

Many teenagers already use general-purpose AI. Your role shifts from gatekeeper to guide.

How to explain it

"Large language models work by predicting the next word, again and again, based on patterns learned from enormous amounts of text. They're very good at producing fluent, convincing writing. But they don't have beliefs or experiences, and they can be confidently wrong."

"That makes them powerful and risky at the same time. Powerful because they can help you research, write, code and create. Risky because they sound sure even when they're wrong, and because people can use them to mislead."

What to emphasise

• AI amplifies what people do with it, for better and worse
• Critical thinking matters more, not less
• The big questions are real: fairness, privacy, misinformation, jobs
• They will help decide how society uses AI

Topics to discuss

AI and careers. Be honest that AI will change many jobs. Skills like judgement, creativity, communication and asking good questions become more valuable.

Misinformation. AI makes convincing fake content cheap to produce. Talk about how that affects news, elections and trust.

Their own use. Ask how they use AI, with curiosity rather than judgement. You may be surprised by how thoughtful they already are.

AI companions. Some apps are designed to act like a friend or partner. Common Sense Media's 2025 research concluded that no one under 18 should use AI companion platforms. Talk about why a product built to keep you talking is different from a real friendship.

Tips That Work at Every Age

Keep it ongoing

This is not a one-off talk. Come back to it as your child grows and as the technology changes.

Use AI together

The best way to teach AI is to use it side by side. Ask questions together, evaluate answers together, and talk about what you notice.

Admit what you do not know

You do not need to be an expert. If you cannot answer, say so and look it up together. That models exactly the curiosity you want them to have.

Lead with curiosity, not fear

AI is a tool. Children who approach it with curiosity and healthy scepticism use it far better than those who are frightened of it or trust it blindly.

Start from their interests

A dinosaur fan and an aspiring artist will meet AI in different ways. Let their passions lead the conversation.

Frequently Asked Questions

At what age should I explain AI to my child?

As soon as they meet it, often around age 4 or 5 through voice assistants or apps. Start with "it's a computer program, not a person" and build from there.

How do I explain AI to a child in one sentence?

"AI is a computer program that learned from lots of examples, so it can answer questions and make things, but it isn't alive and it can make mistakes."

Should I tell my child AI can be wrong?

Yes, at every age. It is the single most protective idea you can give them.

The Bottom Line

Talking to your kids about AI is not about teaching computer science. It is about helping them understand a tool that will shape their lives as students, workers and citizens. Start simple, be honest, stay curious, and keep the conversation going. For the safety side of the picture, see our complete guide to AI safety for children.