Parents' AI Safety Checklist: 15 Rules Every Family Should Know

A parent keeps the list near the study area. When a new app asks for access to every photo, the family pauses at the permissions rule instead of accepting automatically. A short conversation prevents an unnecessary upload.

Which fifteen rules should we remember?

1. Check the real provider. Start from its official website or linked store listing; a logo is not proof.
2. Read the age rules. Class level is not account permission; use adult operation where required.
3. Agree on the task. Know what you want to learn before opening the tool.
4. Keep identity private. No Aadhaar, PEN, APAAR, full address or daily travel routine.
5. Protect account secrets. Never share passwords, OTPs, recovery codes or payment approvals.
6. Avoid private child images. Prefer fictional characters and drawings; consider every person in an image.
7. Limit permissions. Review camera, microphone, location and photo access; grant only what the task needs.
8. Read data controls separately. Training, history, memory, sharing and deletion are different matters.
9. Try before asking. Encourage an independent attempt, then a hint rather than a copied answer.
10. Verify important claims. Open the source and check dates, numbers and context.
11. Follow school instructions. Acknowledge permitted help honestly; do not use AI to bypass an assessment.
12. Keep human relationships central. A chatbot is not a friend, counsellor or emergency responder.
13. Stop for pressure or distress. Secrecy, unsafe content and demands for money need an adult's attention.
14. Protect sleep, play and conversation. End a bounded activity instead of allowing endless use.
15. Report and get help. Tell a trusted adult, use official channels and do not forward harmful child sexual material.

Which rule should we practise first?

Choose one relevant to the family's current activity. For example, read permissions together and explain why a spelling task does not need contacts. UNICEF's privacy guidance supports reviewing permissions and keeping conversations open. AI-S12: Read the source

What should we do after a mistake?

Listen without blame, stop further sharing and secure exposed accounts. Financial cyberfraud needs prompt contact with the bank and 1930, followed by the official reporting process. MHA confirms these Indian mechanisms; a report does not guarantee recovery. AI-S20: Read the source

For immediate physical danger seek emergency help. For suspected child sexual offences contact the appropriate police; a school discussion alone is not a substitute.

How do we keep the checklist useful?

Review it when the child grows, a product changes or a concern appears. Let children explain a rule in their own words and invite them to remind adults too. A checklist works best as a shared habit, not a reason to blame a child after harm.

Where can you check the details?

• AI-S12: Online privacy checklist for parents
• AI-S20: National Cybercrime Response Mechanism

What can you read next?

• 10 Safe AI Prompts Parents Can Try for Their Children's Learning
• How to Choose an Age-Appropriate AI Learning Tool for Your Child
• 10 Things You Should Never Share with AI Chatbots