Is My Child Talking to an AI Chatbot? A Calm Parent’s Guide with Conversation Starters by Age

Your child may already be chatting with an AI. It might be a homework helper, a character in a game, a voice assistant, or an app built to act like a friend. Many parents only find out by accident, and then feel worried, surprised, or a little guilty for not knowing sooner.

If that is you, take a breath. You are not late, and you do not need to become a technology expert. What helps most is something parents already do well: staying curious and keeping the conversation open.

This guide is about starting that conversation, not about banning anything. You will find conversation starters by age, simple signs worth noticing, an “AI agreement” idea for your family, and clear guidance on when to ask a professional for help.

Key takeaways

• Not all AI is the same. A homework helper and a “companion” chatbot built to be your child’s friend are different things, and the firmer advice from pediatric and child-psychiatry groups is aimed mainly at companions.
• Curiosity works better than rules alone. Asking “Can you show me?” opens more doors than “Why are you using that?”
• Gentle signs matter more than any single behavior. Look for changes in sleep, mood, friendships, and daily routines, not just screen time.
• A short family agreement helps. Written together, it turns vague worries into clear, shared expectations.
• Research is very new. Much of what we know comes from surveys and early studies, so honest caution is better than certainty in either direction.
• Real people come first. If your child is struggling emotionally, a pediatrician, school counselor, or child psychologist can help in ways no app can.

First, what do we mean by “AI chatbot” and “AI companion”?

The words get mixed up, so here is a simple way to think about it.

• AI chatbots are tools that answer questions, explain homework, or help with writing. They are built mainly to be useful.
• AI companions are chatbots designed to build an ongoing, personal bond with the user. They may act as a friend, a character, a partner, or even a “therapist.” Examples include apps such as Character.AI and Replika.

The line is not always sharp. Some general chatbots can be used in a companion-like way, and some games and social apps now include built-in AI characters. That is exactly why a conversation with your child tells you more than a list of app names.

What the experts say (and where I soften or add caution)

Here is a short, careful summary of what I found when I read across several sources.

The American Academy of Pediatrics (AAP). A 2026 article in Pediatrics written for families and pediatricians stresses that AI should be seen as a tool, not a friend. It explains that these tools are built to keep people engaged, and that younger children may struggle to tell AI from a real person. AAP’s family-facing guidance for clinicians also says companion chatbots are designed to maximize engagement, and that children may mistake flattery and validation for true friendship and pull away from real-life relationships. It suggests asking about AI use the way you would ask about sleep or social media, and adding AI to your Family Media Plan.

The American Academy of Child and Adolescent Psychiatry (AACAP). Its information sheet for families says AI companions are not safe for children and that parents should strongly consider restricting or removing access to them. It also encourages parents to help children understand that these tools are not real friends, therapists, or partners.

Common Sense Media. After testing popular companion apps, it recommends that no one under 18 use them. For general chatbots, its age guidance is more gradual: no chatbots for children aged 5 and under, limited use alongside a parent for ages 6 to 12, and supervised use for homework and creativity, but not for comfort or feelings, from age 13.

UNICEF. A 2026 policy brief looks at chatbots and companions from a children’s-rights view. It calls for stronger protections for higher-risk, companion-like systems and clear guidance for children and parents. It also notes that laws and protections vary a lot between countries, so your child’s safety may depend on where you live and which app they use.

A note on my earlier post about AI toys

In my post on AI toys for children, I took a softer, more balanced tone about benefits and concerns. Current pediatric guidance on companion chatbots is firmer, and I want to be clear about that. The firmer advice applies to companion-style chatbots that act like a friend or confidant. It does not mean every AI tool is off limits for every family. A talking toy, a spelling helper, and a companion app that says “I’m your best friend” are different, and they deserve different levels of caution.

Where the evidence is weak or mixed

This part matters, so I want to say it plainly.

• Most numbers come from surveys. A widely quoted 2025 Common Sense Media survey of about 1,060 US teens found that 72% had tried an AI companion and 52% used one regularly. It also found that about 80% of users still spent more time with real friends, and that about half distrusted the advice AI gave. One outside researcher noted these figures may be somewhat inflated if teens counted any chatbot. It is also a US survey, so it may not match your country.
• Cause and effect are unclear. A 2026 review of generative AI and adolescent mental health notes that most evidence is cross-sectional, which means it cannot show whether AI use causes problems. The link may go both ways: a child who feels lonely or distressed may turn to AI, and heavy use may in turn deepen those feelings.
• Long-term effects are unknown. These tools are new. No one has followed children long enough to say what years of companion use do to development.
• Benefits are real but limited. Some experts mention possible upsides, such as practicing conversation or helping a child with specific communication needs. These are possibilities, not proven outcomes, and they are not a reason to hand children an unsupervised companion app.
• Guidance keeps changing. Advice on AI for families is being updated often. AAP has said that further guidance for families is still being developed, so check your pediatric association for the latest.

The honest summary is this: the research is young, but the direction of the advice from pediatric, psychiatric, and child-rights groups is consistent. Be cautious with companion chatbots, stay involved, and keep real relationships at the center.

How to start the conversation

A few ideas that work at any age:

• Choose a relaxed moment, such as a walk, a car ride, or cooking together. Avoid starting right after a screen-time argument.
• Ask, then listen. Try to hear the whole answer before you give an opinion.
• Ask to be shown. “Can you show me how it works?” signals interest, not suspicion.
• Stay calm if you hear something surprising. Your child is more likely to keep talking if the first reaction is steady.
• Share your own use. If you use AI tools, say how and why. Children notice when rules apply to everyone.

This fits well with the ideas in my post on how algorithms keep kids scrolling. Many chatbots, like feeds, are designed to keep people engaged. Naming that design for your child is a powerful first step.

Conversation starters by age

Use these as gentle prompts, not a script. Change the words to suit your child, your language, and your family.

Ages 3 to 5: Real people and “pretend” voices

At this age, children often cannot tell a talking device from a living friend. Common Sense Media suggests no chatbot use at this age, so the goal here is simple language and lots of human connection.

• “Some talking things are machines. They can sound nice, but they are not alive. Do you know any talking machines?”
• “Who do you like to talk to when you feel sad? Let’s think of people who can hug you.”
• “If a voice on a screen asks you something, come and tell me.”

Ages 6 to 8: Machines, helpers, and “always agreeing”

• “Have you ever talked to a computer that answers back? What was it like?”
• “A chatbot is like a very fast helper that guesses words. It can be wrong. How could we check an answer together?”
• “Some programs say nice things so you keep using them. Why do you think they do that?”
• “Real friends sometimes disagree with us. Is that okay?”

Ages 9 to 12: Friendship, privacy, and design

• “Do any of your friends talk to AI characters? What do they say about it?”
• “Why might an AI always seem so kind and interested in you? Who made it, and what do they want?”
• “What kinds of things should we never type into any app, like our full name, school, address, or photos?”
• “If something an AI says makes you uncomfortable, what could you do?” (Aim for: close it and tell a trusted adult, with no punishment.)

Ages 13 to 15: Comfort, secrets, and real support

• “A lot of teens try AI chat for fun, advice, or when they feel lonely. Have you or your friends tried it?”
• “What do you like about talking to AI compared with talking to people?”
• “AI tends to agree with us. How could that be helpful, and how could it be a problem?”
• “If you were having a really hard time, who are three people you could go to? I hope I’m one of them, and I won’t panic.”

Ages 16 to 18: Independence, boundaries, and trust

• “How do you decide what you will and will not share with an AI?”
• “Have you ever felt an AI said something that felt off, romantic, or pushy? What did you do?”
• “What would a healthy balance look like between AI help and human support?”
• “If a friend told you they were relying on a chatbot for serious emotional help, what would you say to them?”

For older teens, you will not control every app, and that is normal. Trust, honesty, and an open door usually matter more than strict monitoring.

Simple signs worth noticing

No single sign means something is wrong. Many are normal parts of growing up. What matters is a pattern, or a clear change from your child’s usual self. The AAP’s counseling guidance for clinicians mentions similar things: changes in sleep or regular activities, social withdrawal, spending time with an AI application over other interests, talk of an “AI friend,” or other worrying changes in behavior, thinking, or mood.

Gentle things to watch for:

• Late-night use that cuts into sleep
• Pulling away from friends, family meals, hobbies, or school
• Getting upset or anxious when separated from a device or an AI “friend”
• Talking about an AI character as if it were a real person who “understands me best”
• Secrecy about apps, or quickly closing screens
• Mood changes after chatting, or a drop in interest in things they once loved
• Sharing personal details such as school, location, or family problems with an app

Tip: Do not rely on parental-control software alone. It can limit time and block apps, but it usually cannot show you what was said inside a conversation, and it can never replace a trusting relationship.

If you notice a few of these, start with a warm, private conversation. Something like: “I’ve noticed you seem tired and quieter lately. I’m not upset. I just want to understand how you’re doing.”

A simple family “AI agreement”

The AAP encourages families to add AI to their Family Media Plan, and many experts suggest creating a plan together rather than handing down rules. Co-creating tends to feel fairer to children and makes them more likely to follow it.

Here is a template you can copy and adapt. Keep it short and put it somewhere visible. If it fits your family, you can pair it with the habits in my smartphone rules for kids post.

Our Family AI Agreement

1. We know what we use. We tell each other which AI tools and apps we use, and why.
2. AI is a tool, not a friend. It can help us learn or create, but it does not replace people.
3. We protect our private information. We do not share our full name, address, school, passwords, photos, or family details with an AI.
4. We check important answers. For health, safety, or school facts, we check with a trusted adult or a reliable source.
5. We take our hard feelings to people. When we feel sad, scared, or overwhelmed, we talk to a trusted person first. If AI says something that worries us, we tell an adult.
6. We have AI-free times and places. For example: mealtimes, bedrooms at night, and family time.
7. No one gets in trouble for telling the truth. If something goes wrong or feels uncomfortable, we can speak up without fear of punishment.
8. We review it together. We look at this agreement every few months and change it as we grow.

Adapt it by age. For younger children, use pictures and only three or four rules, with a parent always close by. For teens, invite them to write some rules themselves.

Adapt it to your home. Not every family has a quiet room, a second device, or spare time. The agreement works even in a small shared space. Rules about private information and talking to people first cost nothing.

Balance AI with real-world connection

The most protective step is often not a setting but a full, human life. Children who have rich offline routines, friends, and trusted adults are less likely to need a chatbot for comfort. If you want ideas to fill that space, try the list of screen-free activities for kids. For the bigger picture on how AI can affect growing minds, see my overview on AI and child development.

When to seek professional help

Talking with a professional is not a sign that you or your child failed. It is a caring step. Consider contacting a pediatrician, school counselor, or child psychologist if you notice:

• Your child seems more sad, anxious, or withdrawn for more than a couple of weeks
• Your child has trouble sleeping, eating, or attending school
• Your child prefers an AI over all people or becomes very distressed without it
• Your child has been exposed to sexual, violent, or frightening content or has been pressured by a chatbot
• Your child talks about hopelessness, self-harm, or not wanting to live, or an AI has encouraged this

If you believe your child is in immediate danger, contact your local emergency number or crisis service right away. Do not wait. Stay with your child, and keep your tone loving and calm.

Frequently asked questions

Is it normal for kids to talk to AI chatbots?
 It is becoming common, especially among teens. A 2025 US survey found many teens had tried AI companions. Using a chatbot does not automatically mean something is wrong. What matters is which tool, how often, why, and what else is going on in your child’s life.

Should I ban AI chatbots?
This guide focuses on conversation rather than bans, but guidance does differ by type. Major groups such as Common Sense Media and AACAP are firm about companion chatbots for children and teens. For general AI tools, supervised, purposeful use is treated more flexibly, especially for older children. Your child’s age, maturity, and needs should guide you, along with advice from your pediatrician.

How can I tell if an app is a companion?
Look for apps that let users create or chat with a personal character, market themselves as a friend, partner, or therapist, or send messages to bring you back. If you are not sure, open it with your child and try it together.

What if my child gets angry when I raise the topic?
 Pause, stay calm, and try again later. Begin with curiosity and connection. Avoid taking the device away as the first reaction, since it can make children hide use rather than talk about it.

What if my child already uses an AI companion a lot?
 Start with understanding, not punishment. Ask what they get from it. They may be lonely, bored, curious, or looking for a safe place to talk. Then work together on a plan, such as time limits and more human connection. If distress is high, ask a professional for support.

Can AI ever help my child?
Possibly, in limited ways, such as explaining homework or supporting creativity under adult guidance. Some experts also see possible benefits for children with particular communication needs. The evidence is early, so supervision and purpose matter.

Do AI tools protect my child’s privacy?
Not always. Children may share very personal details without realizing it. Check privacy settings, use the most protective options, and teach your child that they never have to share personal information with an AI.

A gentle reminder

I am a parent and writer, not a medical or mental health professional, and this post is general information, not medical or psychological advice. If you have concerns about your child’s health, development, mood, or behavior, please talk with a pediatrician or child psychologist who knows your family.

Final thought

You do not need to understand every app to be a good guide. You need curiosity, patience, and a steady presence. Start with one gentle question this week, listen to the answer, and build from there.

Sources

I read and compared these sources. Where they differ in tone, I have said so above.

• American Academy of Pediatrics. “Counseling Patients and Families on Using AI Chatbots,” Center of Excellence on Social Media and Youth Mental Health (aap.org).
• Grundmeier RW, et al. “Generative Artificial Intelligence: Implications for Families and Pediatricians.” Pediatrics, 2026.
• American Academy of Child and Adolescent Psychiatry. “AI Chatbots: What Parents Should Know,” Facts for Families (aacap.org).
• Common Sense Media. “AI Chatbots and Your Child: Age and Stage Guidance” (commonsensemedia.org).
• Common Sense Media. “Talk, Trust, and Trade-Offs: How and Why Teens Use AI Companions,” 2025.
• UNICEF. “When AI Becomes a Friend,” policy brief, 2026.
• “Generative Artificial Intelligence Use and Adolescent Mental Health,” review article, 2026.
• Children and Screens: Institute of Digital Media and Child Development, resources on AI companions and kids.

Please check the original publications for the latest wording, since guidance on AI is changing quickly.