Chatbots and AI "friends"
Talking to an AI is now part of everyday life for many children. Used well, it is a genuinely useful tool. The trick is understanding where the real risks are, so you can set sensible boundaries without banning it outright.
The ones children actually use
Most set a minimum age of 13, and some ask for a parent's permission for under-18s. Character.AI now restricts its chat to over-18s. Ages and rules change fast, so always check the current terms, and see the AI tools reference for the detail.
ChatGPT (OpenAI)
A general chatbot for questions, writing and homework help.
Google Gemini
Google's chatbot, built into search and many Android phones.
Snapchat My AI
A chatbot inside Snapchat, so many children already have it to hand.
Character.AI
Lets users chat with AI "characters" that feel like real people. Since late 2025, its chat is restricted to over-18s.
Meta AI
Built into WhatsApp, Instagram and Messenger.
The real risks, in plain terms
Confident, but sometimes wrong
Chatbots can state wrong information as if it were fact, and sometimes invent things entirely. Children can take an answer at face value, especially for homework.
Emotional attachment
Companion apps are built to feel like a friend who is always available and always agrees. Some children can lean on them for support or form a strong attachment, which is not a substitute for real relationships.
Grown-up content
Filters are not perfect. Depending on the app and how it is used, a child may come across content or conversations that are not suitable for their age.
Data and privacy
What your child types can be stored and used to train the AI. Children may share personal details, photos or worries without realising where that information goes.
What healthy use looks like
- Used as a tool to learn or create, not as a secret friend to confide in.
- Answers are double-checked, not simply trusted or copied.
- Used in a shared family space rather than alone behind a closed door.
- Your child knows they can tell you if a chatbot says something odd or upsetting.
Warning signs to watch for
- Choosing to talk to an AI companion instead of friends or family.
- Getting upset or secretive about time spent on a chatbot app.
- Treating the AI as their main source of advice or emotional support.
- Sharing personal information, photos or feelings they would not share elsewhere.
If you notice these, it is worth a gentle conversation. The parent playbook has starters and support links.
Practical steps that help
- Set up any account with your child's real age, so age protections apply.
- Look for a teen or parental setting in the app and turn on the strongest option.
- Keep chatbot use in shared spaces, not alone in the bedroom.
- Agree together what it is fine to use AI for, and what it is not.
Read the parent playbook
Age-by-age guidance, conversation starters and a family agreement.
Last reviewed: July 2026.