Chatbots can answer questions, help with ideas and hold a conversation. Some products emphasise tasks; others are designed around companionship or characters. The same conversational style can conceal very different purposes, settings and risks.
Assistants and companions
Ask your child what they use the chatbot for. Helping explain a maths question, writing a story and treating a character as a close confidant involve different expectations. The distinction helps you choose boundaries without assuming all conversation with AI is the same.
A friendly response can feel reassuring. Take that experience seriously, while explaining that a chatbot should not become the only place a child turns for difficult personal decisions. Make sure trusted people remain accessible.
Potential benefits and real limits
A chatbot may support practice, creativity or an initial explanation. Encourage your child to compare its answer with what they already know, ask how to check it and explain the idea back in their own words. That makes the activity about learning rather than accepting an answer.
Outputs can be inaccurate, inappropriate or unsuitable for a child’s situation. A convincing answer about health, safety or a personal crisis needs an appropriate human source. A disclaimer or reassuring tone does not make a system qualified to handle that responsibility.
Avoid entering passwords, private family concerns, identifying school details or sensitive images. Check the product’s actual data controls and terms; do not assume that a private-looking chat remains only on your device.
Check the actual product and current terms
Write down the service name, account type and where your child accesses it. Look for the official minimum-age and parental-permission terms, safety controls, privacy explanation and reporting route. Check web access as well as app access.
Do not substitute your judgment of a child’s maturity for a service’s eligibility requirements. If a tool is not permitted for their age, choose an appropriate activity instead. A school-approved classroom use also does not automatically authorise independent personal use.
This guide does not rate products or publish remembered feature lists. A product-specific recommendation needs current owner documentation and a review of the actual use case.
Agree family boundaries
Discuss when the chatbot is useful and when to bring a person into the conversation. Examples include a distressing answer, sexual content, pressure to keep secrets, spending requests or important advice that contradicts a professional.
Agree how a child can show you a concerning text response without sharing illegal or private images. Explain oversight in advance and make it proportionate. Secret monitoring can damage trust; transparent expectations make it easier to ask for help.
The family agreement includes new apps, personal information and help-seeking. Review the agreement after a change in the product or your child’s needs.
If use is becoming distressing
Look at its effect on sleep, school, relationships and the child’s ability to stop or seek support. Avoid diagnosing dependence from screen time alone. Ask what the chatbot provides and what your child fears losing if the use changes.
If distress is significant or worsening, seek appropriate professional support. If your child is in immediate danger, use emergency help. A chatbot conversation must not replace urgent assistance or ongoing care.
If a response is unsafe
Help your child stop the interaction and use the product’s genuine reporting process. Keep safe text details if useful, without copying sexual imagery of a child. If a human is using the service to exploit or threaten them, use the relevant police or child-protection route.
Maintain a calm path back to real support: “You can tell me if something goes wrong, even with a tool we agreed you could use.” NSPCC provides further parent guidance on AI conversations. (NSPCC talking to children about AI)