Help your child learn with AI: a parent's guide

Help your child use AI by checking one answer together and asking them to explain what they learned afterwards. The routine below draws on research and official safety guidance. We have not tested the routine as a whole.

Screenshots show public guidance and a source-checking example.

1. Agree on the tool and the rules

1. Ask which AI tools your child uses and why, including chatbots used for company.
2. Check the provider's current minimum age, parent permission requirements and the school's assignment rules. Do not enter a false age or give a child your login to bypass restrictions.
3. Review privacy and parental settings together. Explain their limits and what you can see.
4. Agree when and where the tool may be used, what information stays private and who the child can ask for help.

Keep names, school details, location, faces and private problems out of prompts. If your child is ineligible, prepare and check an activity yourself, where your own use is permitted. Use the finished material offline together.

Interviews with seven US teenagers and 13 parents found differences in their understanding of AI use and risks, including companionship. This small study cannot establish how common these experiences are or prove that a parenting method prevents harm. Yu et al., 2025; open author version (opens in a new tab)

2. Choose one task to do together

Pick a topic your child is learning. Ask what they think before opening AI. Choose an end point, such as explaining day and night in their own words.

For eligible users, a possible prompt is: "Ask me one question about why we have day and night. Wait for my answer, then give feedback. Keep it brief."

Check the feedback together. Correct mistakes using a reliable source, or finish without AI.

A five-week study observed 18 children aged 5 to 11 and 16 parents learning about AI together. Parents sometimes taught, collaborated or learned from their children. The study covered image recognition, voice assistants and offline activities. It did not test today's generative chatbots or establish that children can use them safely without supervision. Druga, Christoph and Ko, 2022 (opens in a new tab)

For an offline activity, use a ball and torch to explore which side receives light, then read an explanation together.

3. Check a fact together

1. Pick one claim in the answer. Ask: "How could we check that?"
2. Open NASA's What Is Earth? (opens in a new tab). Select Why Do We Have Day and Night? in Contents.
3. Compare the explanation with the claim. NASA describes how Earth's spin brings most places through day and night, with exceptions near the poles.
4. Ask what the source confirms or changes.

Asking the chatbot whether it is sure does not replace checking the source.

Enlarge image

NASA explanation of day and night with the polar daylight exception

×

NASA’s What Is Earth? explanation, including the polar exception. Cropped source-checking screenshot.

Open full-size image

4. Close AI and ask for an explanation

Ask your child to explain the idea without the chatbot, or try a related question on paper. Keep any accessibility support they need. If they cannot explain it, return to the book, teacher or worked example.

In a randomized study at one high school in Turkey, a basic GPT-4 tutor improved assisted practice scores but reduced subsequent maths exam scores without AI. A specially designed tutor avoided that observed penalty. The study does not establish effects on young children's reading. Bastani et al., 2025 (opens in a new tab)

Check what your child understands. Finishing homework faster is not enough evidence of learning.

5. Practise stopping and asking for help

Explain that an AI reply is not a trusted adult's advice. Use a tool intended for the child's learning task and avoid apps marketed as romantic companions.

Practise responding to upsetting content or requests for secrets: close the chat and tell a trusted adult. Say they can ask for help even if they broke a rule.

eSafety recommends supervision and conversations alongside parental controls. Controls cannot prevent every harmful interaction. Open its parental controls guide (opens in a new tab) and select Building online safety habits for the guidance shown below.

Enlarge image

eSafety guidance explaining that parental controls cannot prevent every online risk

×

eSafety’s Building online safety habits guidance. Text © Commonwealth of Australia, CC BY 4.0. Cropped screenshot.

Open full-size image

If AI use becomes upsetting or difficult to stop

Pause the activity and talk with your child about what happened. Use the service's reporting option for harmful content. If distress continues, seek support from the school's wellbeing staff or a qualified local professional. For immediate danger, contact your local emergency service.

Papers and guidance

• Yu et al. (2025), Exploring Parent-Child Perceptions on Safety in Generative AI: Concerns, Mitigation Strategies, and Design Implications. IEEE Symposium on Security and Privacy. Published record (opens in a new tab); open author version, revised 2024 (opens in a new tab).
• Druga, Christoph and Ko (2022), Family as a Third Space for AI Literacies: How do children and parents learn about AI together? CHI. DOI (opens in a new tab); open author paper (opens in a new tab).
• Bastani et al. (2025), Generative AI without guardrails can harm learning: Evidence from high school mathematics. PNAS. Paper and DOI (opens in a new tab).

Official guidance: eSafety's parental controls guide (opens in a new tab) and AI chatbot advisory (opens in a new tab).

Screenshot text: eSafety Commissioner, CC BY 4.0 (opens in a new tab). Screenshots are cropped; no endorsement is implied.