A practical guide for parents on ChatGPT — what it is, how children use it, and how to set appropriate boundaries around AI chatbot use.
ChatGPT is a popular AI chatbot from OpenAI that answers questions, writes text and helps with tasks through natural conversation.
Official age
13+
We recommend
14+
Developer
OpenAI
Risks
3
ChatGPT is an AI chatbot developed by OpenAI and one of the most widely used AI tools in the world. It can hold extended conversations, answer questions, write essays, summarise text, generate code, and assist with creative projects. Because it is accessible via a free web browser as well as dedicated apps, children often discover and use it without parental knowledge. Its ability to produce fluent, plausible-sounding text makes it both a powerful study tool and a potential source of misinformation.
Children most commonly use ChatGPT to get homework help, draft essays, or research topics. Many also use it for creative writing, role-play scenarios, or simply out of curiosity. Older teenagers may use it for coding assistance or to explore ideas they feel uncomfortable raising with adults. Some children share personal details during conversations without considering that their input may be used to improve the model.
Memory / Chat History
Location: Settings → Personalisation → Memory / Improve the model for everyone
Set to: Disabled
By default, ChatGPT can save chat history and use it to train future models. Turning this off limits data retention and reduces the chance that personal details shared in conversations are stored long-term.
Shared Links
Location: Conversation → Share → Manage shared links
Set to: Review regularly
ChatGPT allows conversations to be shared via a public link. Discuss with your child that shared chats are accessible to anyone with the link and should never contain personal information.
Data Controls
Location: Settings → Data Controls
Set to: Review and restrict as appropriate
OpenAI provides options to delete conversation history and opt out of model training. Review these controls and ensure your child understands their data does not simply disappear after a chat ends.
Review your child's recent ChatGPT conversations together and discuss what kinds of questions are appropriate to ask an AI
Time: 20 minutes
Explain AI hallucinations — that ChatGPT can state incorrect information confidently — and the importance of checking facts from reliable sources
Time: 15 minutes
Agree clear boundaries on when ChatGPT use is acceptable (e.g. for brainstorming but not for submitting as original work) and discuss academic integrity
Time: 15 minutes
In immediate danger: call 999. For non-emergency police matters, call 101.
Concerned about a child but it's not an emergency? NSPCC helpline 0808 800 5000. Childline for young people 0800 1111.
Myth: ChatGPT is like Google — the answers are just facts.
Fact: ChatGPT generates plausible-sounding text; it doesn't look facts up. It can be confidently and fluently wrong, and a child has no way to tell from the answer itself. Anything that matters — homework, health, money — should be treated as a starting point to check elsewhere, not an answer.
Myth: There must be a kids' mode or parental controls built in.
Fact: There is no robust age verification and nothing resembling the parental controls of a games console. OpenAI's terms require users to be 13 or over, with parental consent for under-18s, but nothing enforces it — the real safeguards are your involvement and the settings you review together.
Myth: It's only a homework tool, so the social-media rules don't apply.
Fact: Children also use ChatGPT for role-play, emotional conversations and questions they'd never ask an adult — and they share personal details without realising chats can be stored and used to train future models. It deserves the same open, curious conversations as any messaging app.
Below ChatGPT's minimum age of 13, so no account of their own. If they're curious about AI, explore it together on your account — ask it something you both know about and let them spot the mistakes, which teaches the most important lesson early: it makes things up.
Still under the minimum for most of this band. At 13, OpenAI requires parental consent for teenagers, so the setup is legitimately yours to be part of: create the account together, turn off memory and model training in the data controls, and agree what it's for — ideas and explanations, not finished work.
The two conversations that matter are honesty and accuracy: agree a 'brainstorm, don't submit' rule that matches their school's AI policy, and get them into the habit of checking anything factual against a real source. Remind them never to paste in names, addresses or anything private — theirs or anyone else's.
Watch for over-reliance rather than misuse: using it as a late-night confidant or letting it make decisions for them. Encourage them to keep verifying important answers, and to remember that these chats sit on a company's servers — a conversation with ChatGPT is not a private diary.