Skip to content
Bot Tamers

How to teach your kids about AI at home (ages 8 to 12)

By the Bot Tamers team. Last updated .

You can teach your kids about AI at home without being an expert. Start with how AI works, with safety and fairness built in from the start, then build up to using AI well in real life. Spend about 15 minutes at a time on one short activity together, then ask your child to explain the idea back to you.

This guide gives you that path, five short activities, what changes between ages 8 and 12, and a simple routine. Each activity takes a kitchen table, a phone or laptop, and you.

You don't need to be an AI expert

Many parents hold off because they feel they don't know enough. Ying Xu, who studies AI and learning at Harvard, told UNICEF that it helps to "move away from the idea that parents must always act as experts or supervisors." Adults and children are often learning about AI at a similar pace, so she suggests approaching it "as co-learners."

In practice, that means saying "I don't know, let's find out" out loud, and thinking out loud when you use AI yourself. The NSPCC suggests the same thing: show your child how you question an AI answer, and talk about accuracy, sources and fairness as you go.

Start with the AI your child already meets

Before any lesson, find out where AI already shows up in your child's day. The NSPCC suggests checking which apps and services your child uses and whether they include AI features. Its examples include chatbots, tools that make images, stories or music, deepfakes, and AI summaries on search results.

Make it a game: spend five minutes listing every place you both think AI might be at home, then talk about the ones you are not sure of. Our guide on how to talk to your child about AI has more ways to open the topic.

The path: how AI works first, then real life

Safety and fairness are not a separate topic you save for the end. They come straight out of how AI works, so teach them along the way. A child who knows how AI makes an answer finds it easier to see why it can be unfair or wrong, and why it should not be trusted like a person. Our guide to AI literacy for kids explains the skills behind each part in more depth.

Start with how AI works: it learns patterns and makes guesses

Start here. AI is a tool made by people. It learns patterns from lots of examples and uses them to guess what comes next. As Ying Xu puts it for UNICEF, AI tools follow instructions, recognize patterns or respond to our voices, but they do not "think" or "feel" the way people do.

Try at home: the next word game. Say the start of a phrase your child knows, like "peanut butter and," and let them finish it. Then try one with many possible endings, like "My favorite animal is." Explain that a chatbot does something like this, one word at a time, based on the text it learned from. Northwestern University's free AI Unplugged collection has a classroom version, Large Language MadLibs, in which students "simulate how large language models generate text using prediction."

Build in safety and fairness from the start

Fairness comes straight out of the first idea: if AI learns from examples, then what it learns depends on which examples people gave it. Add three ideas as you go. What it learns can be unfair. What you type can be saved. And it is a tool, not a friend.

Try at home: the one-sided robot. Tell your child about a pretend robot that has only ever been shown pictures of big brown dogs, each one labeled "dog." Ask what it might say when it sees a small white dog. Then ask who could be left out if a real AI only learned from some kinds of people, and how you would fix it. Thinking out loud about fairness like this is one of the habits the NSPCC suggests.

Then cover safety. The NSPCC says to explain that AI tools are not trusted adults, that your child should not share personal information with them, and that they should come to you or another adult if they are worried. Agree on two family rules: we never type our full name, school, address or passwords into an AI tool, and if an AI says something strange or scary, we stop and tell a grown-up. For more, see our guide on kids and AI chatbots.

Then build up to using AI well in real life: ask clearly, check it, and do your own thinking

Now put it to work on the kind of things your child really meets. Show how the question shapes the answer, and that AI should help your child think, not do the thinking for them. Ying Xu points to a real worry here: some children start to hand their learning over to AI, which can take away the "productive struggle" that long-term learning needs.

Try at home: the better question. On your own account and your own device, ask a chatbot a vague question your child cares about, like "Tell me about sharks." Then ask a clearer one together, like "Explain how sharks find food, for a 9 year old, in five sentences." Compare the two answers. Then close the laptop and ask your child to tell you one thing in their own words. UNICEF's advice is to look at the answer together and ask which parts seem helpful, which seem confusing or not quite right, and how the tool might have made it.

Try at home: catch the mistake. AI answers can look right and still contain errors, so a confident answer is not the same as a correct one. The NSPCC suggests simple questions like "Do you think this answer makes sense?" Ask an AI tool about something your child knows inside out, like a game, a sport or a favorite book, and count the mistakes together. Kids enjoy being the expert, and it shows them why checking matters. Our guide on how to teach kids to fact-check AI shows how to check an answer against a trusted source, step by step.

Try at home: real or made? Look at a few pictures together, some real and some made by AI, and ask your child how they would tell. Talk about where each one came from and whether you could check it somewhere else. Our guides on explaining deepfakes to kids and AI voice scams go further.

What changes from age 8 to age 12

The path stays the same. What changes is who types and how deep you go.

  • Ages 8 to 10: you drive. Use your own account on a shared screen, and do the typing while your child asks the questions. Keep each activity short and playful, and lean on games that need no screen at all, like the next word game and the one-sided robot.
  • Ages 11 to 12: they type, you sit next to them. Some 9 and 10 year olds may be ready for this too. Let your child write the questions, compare two answers to the same question, and talk about when AI is fair to use for homework and when it is not.

Before your child uses any AI app on their own, check its age rules and read our guide on whether ChatGPT is safe for kids.

A simple routine that sticks

One activity at a time is enough. Pick a regular time, spend about 15 minutes on it, and ask one question about it a few days later, at dinner or in the car. Go in order: the next word game, then the one-sided robot, then the three real-life activities. When you have done them all, start again and go a little deeper.

If you want ready-made materials, Day of AI and Common Sense Media made a free AI Literacy Family Toolkit with short videos, conversation cards and simple, hands-on activities to do together at home. If you homeschool, our guide to choosing an AI curriculum for homeschool has a fuller plan.

How to tell it is working

You will see it in small moments. Your child can explain what AI is in their own words. They ask who might be left out of an AI's examples. They check an answer without being asked, and they pause before typing anything personal. If, instead, you notice long stretches of chatting with an AI, secrecy, or AI crowding out sleep, schoolwork or friends, Ying Xu suggests starting with gentle, open questions rather than criticism.

Questions parents ask

How do I teach my kids about AI at home?

Start with how AI works: it learns patterns from examples and makes guesses. Build in safety and fairness from the start: what is fair, what to keep private, and why AI is a tool, not a friend. Then build up to using AI well in real life, like asking clearly and checking answers. Spend about 15 minutes at a time on one activity.

What should an 8 year old learn about AI compared with a 12 year old?

The same path, at a different depth. At 8 to 10, you type on a shared screen and use simple games, like guessing the next word. At 11 to 12, your child can type while you sit with them, compare two AI answers, and talk about when using AI for homework is fair.

How much time does it take to teach kids about AI?

Less than most parents expect. About 15 minutes at a time on one activity, plus a question at dinner, is a good start. Common Sense Education's AI literacy lessons for grades K to 12 are listed at 20 minutes or less. Short, regular practice matters more than long sessions.

Should my child use AI with me or on their own?

With you, at first. UNICEF suggests exploring AI together, for example typing your child's question into a chatbot and looking at the answer side by side. For ages 8 to 10, you do the typing. For 11 and 12, they can type while you sit next to them. Check each app's age rules too.

How do I know my child is actually learning about AI?

Ask them to teach you. A child who has learned it can explain in their own words how AI makes a guess, say why an AI can be unfair, check an AI answer against a trusted source, and tell you what they would never type into an app. Watch for these in everyday moments.

Where Bot Tamers fits

If you would rather have the order and the next step planned for you, we make Bot Tamers, a paid course for kids 8 to 12 in two tracks. It follows the same path: kids start with how AI works, with safety and fairness built in from the start, then build up to using AI well in real life. Ages 8 to 10 have 7 lessons. Ages 11 to 12 have 7 units plus a final project. On the 8 to 10 track, lessons are about 10 minutes, with read-aloud.

Lessons teach how to check an AI answer, and what never to share with an app, and, on the 11 to 12 track, the difference between using AI as a shortcut and using it like a tutor. The parent dashboard shows a short summary of each finished lesson with a question to ask at dinner.

Bot Tamers has no live chatbot, and nothing about your child goes to an AI company. Accounts are for parents only. Kids never get their own login or email; a parent sets up each child's profile. It has a 7-day free trial, then costs $19.99 a month or $199.99 a year. The iOS app is coming very soon. In the iPhone and iPad app, kids never see a price. Plans and payment sit behind a parent check.

See every lesson on the curriculum page, read how we handle your child's data on our safety page, or create a parent account.

Sources

More questions parents ask