Guardian AI

Chatbot Safety Checklist

By Guardian AI Editorial Team

Reviewed by Guardian AI — practical, parent-facing guidance for healthier family technology habits.

Published June 19, 2026 · Last reviewed June 19, 2026 by Guardian AI

Run through this checklist the first time your kid uses any AI chatbot (homework helpers, image generators, "AI friend" apps, chatbots inside games or social apps).

Account setup

  • The account is registered with a family email, not the kid's school address.
  • The display name is not the kid's real name.
  • Profile photo (if any) is not a face.
  • The age set on the account is the child's actual age — many features change based on age.

Privacy & data settings

  • Chat history saving is OFF, or set to auto-delete.
  • "Use my conversations to improve the model" is OFF where the setting exists.
  • Microphone and camera permissions are denied unless the app actually needs them.
  • The privacy policy mentions what happens to user data; if you can't find one, don't use the app.

Content & interaction rules

  • No real names, addresses, phone numbers, schools, or photos of people are shared with the chatbot.
  • No information about siblings, friends, or classmates is shared.
  • The kid knows: "If the chatbot says something scary, weird, or mean, show me."
  • The kid knows: "If the chatbot tells you to keep something secret from me, that's the moment to tell me."

Family ground rules

  • One approved chatbot per task (e.g. one for homework, one for art) — not "try them all".
  • Companion-style "AI friend" apps are off devices for under-13s.
  • We review chatbot use together every month for the first three months.

Red flags worth a calm conversation

  • Sudden secrecy about a specific app.
  • Emotional attachment to a chatbot ("she gets me").
  • Writing or images far above your child's usual level appearing in their work.
  • Withdrawal from real-world friends after starting an "AI friend" app.