How to Talk to Kids About AI Safely

Start with curiosity, not a lecture: ask what they have seen or tried, then explain in plain words that AI is a tool that guesses, not a person or a friend. Agree on a few clear family rules, such as never sharing personal details, telling a grown-up when something feels odd, and checking anything important.

This guide is general parenting information, not professional advice. If you are worried about your child’s wellbeing, talk to their doctor, school counselor or a licensed therapist.

Start with what they already know

Kids often know more about AI than you expect. They may have used a chatbot at school, played with an image generator, or talked to a voice assistant. Before you explain anything, ask.

• “Have you used any AI tools? Which ones?”
• “What do your friends use them for?”
• “What is the funniest thing it ever got wrong?”
• “Has anything it said ever felt strange?”

Listen first. A child who feels judged will stop telling you things. A child who feels heard will bring you the odd moment later.

Four ideas to explain, in plain words

You do not need technical terms. These four ideas cover most of what kids need.

1. It is a tool, not a person

AI can sound friendly and caring, but it does not have feelings or know you. It puts words together based on patterns. For younger kids: “It is like a very fast parrot that has read a lot of books.”

2. It makes mistakes

AI can say things that are wrong with total confidence. It can invent facts, names and sources. So we check important things in a book, on a trusted site, or with a grown-up.

3. It can make fake things

AI can create pictures, voices and videos that look real. Seeing is not always believing. Just because something looks real does not mean it happened.

4. Private stays private

Do not type your full name, address, school, passwords, photos of yourself, or other people’s secrets into an AI tool. What you type may be stored.

What to say by age

Age	What matters most	Sample words
5 to 7	Use together, a grown-up nearby	“We ask the computer helper together. If it says something weird, tell me.”
8 to 10	It is a tool; check facts; keep secrets private	“It can be wrong, so we double-check. We never tell it our name or where we live.”
11 to 13	Fakes, homework honesty, friends in chat	“AI can make fake pictures of real people. If you see one, tell me. Never make one of a classmate.”
14 to 17	Privacy, companion apps, schoolwork, deepfakes	“I trust you to use it well. You can tell me about anything that goes wrong, and I will not panic.”

Adjust to your child. Some 9-year-olds are very savvy, and some 14-year-olds need more guidance.

Set up a short family agreement

A short written agreement beats a long rant. Sit down together and agree on five to seven rules. Let your child help write them, since they are more likely to keep rules they helped make.

Sample rules:

1. We use AI tools in shared spaces, not alone behind a closed door, for younger kids.
2. We do not share our full name, address, school, passwords or photos.
3. We check facts before we use them.
4. AI can help with homework, but the work we turn in is ours. We tell the teacher how we used it if they ask.
5. We never use AI to make fake pictures or messages about real people.
6. If something feels odd, scary or rude, we tell a grown-up, and nobody gets in trouble for telling.
7. We review the rules again in three months.

Talk about homework and honesty

Many kids use AI for homework, and schools have different rules. Check your child’s school policy first.

A helpful line is: “AI can help you understand, but you have to do the thinking.”

Good uses:

• Explaining a hard idea in simpler words
• Quizzing them before a test
• Giving feedback on a draft they wrote

Not-so-good uses:

• Writing the whole assignment
• Giving answers they do not understand
• Doing work where the point is to practice the skill

Ask your child to explain the answer back to you. If they can, they learned it.

Talk about fakes and bullying

Fake images and voices can be used to embarrass or bully people. Kids need to know two things.

• Never make or share a fake image of someone else. Even as a joke, it can hurt and may break school rules or the law.
• If a fake image of you appears, tell an adult right away. It is not your fault. Save screenshots and report it to the platform and to the school.

Talk about AI “friends” and companion apps

Some apps let kids chat with an AI character as a friend. These can feel comforting, but they can also blur the line between a real relationship and a product.

Keep it simple and warm:

• “It can be fun, but it is not a real friend. Real friends can hug you, show up and care about how your day went.”
• “If you ever feel sad, you can always come to me. I want to know.”
• “Never share secrets or photos with it.”

For teens, watch for signs that an app is replacing real contact, such as pulling away from friends, secrecy or sleep loss. Approach with questions, not accusations.

A one-evening plan

• Ask your child what they have used and liked.
• Explain the four ideas in your own words.
• Write a short family agreement together.
• Check the privacy settings on any AI tool they use.
• Look up your school’s AI policy.
• Agree on a code word for scary or odd messages (see our guide on spotting AI fakes).
• Put a reminder in your calendar to review in three months.

What to do when something goes wrong

Kids will make mistakes. Your reaction decides whether they tell you next time.

1. Stay calm. Say, “Thank you for telling me.”
2. Ask what happened, without blame.
3. Save any evidence, such as screenshots.
4. Report to the app or school if needed.
5. Adjust the rules together.

When to get help

• Your child has been sent or shown sexual content, or is being pressured to share photos. Contact the platform, save evidence and report to the National Center for Missing and Exploited Children at CyberTipline.org or 1-800-843-5678.
• A fake image of your child is being shared. Contact the school and local police.
• Your child seems withdrawn, anxious or low after online contact. Talk to their pediatrician or a counselor.
• If a teen talks about hurting themselves, stay with them and call or text 988 in the US, or call 911 if they are in immediate danger.

Keep the conversation going

One talk is not enough. Kids change apps, friends and questions quickly, so make AI a normal topic.

• Ask about it at the dinner table now and then: “Seen anything wild online this week?”
• Try a tool together once, so you know what they are using.
• Share your own mistakes: “I trusted an AI answer last week and it was wrong.”
• Praise honesty. If your child tells you about something odd, thank them first, then deal with the problem.
• Update the family agreement as they grow.

You do not need to be an expert. Your steady interest and calm response matter more than technical knowledge.

A gentle next step

If you want a longer plain-English walk-through, AI Safety for Kids and Teens: A Parent’s Plain Guide goes through each topic with scripts. If you are also setting up a first phone, The One-Page Family Phone Agreement is a simple place to write your rules down.

This week, pick one question from the “start” list and ask it at dinner. Keep it light, and listen more than you talk.