Five principles
Simple ideas children can remember. Powerful habits for life.
Use these principles whenever a child encounters an AI answer, image, companion, or connected toy. Open each one for the explanation and a question to ask together.
Back to the AI Family Hub
Stay in charge of AI
Five principles for safer, smarter use.
These are not technical rules. They are durable questions and habits that help children pause, check, and decide.
01AI is guessing—really fast.It predicts patterns. It does not understand what it says.
Generative AI learned from enormous collections of human-made material. It predicts what belongs next, which can look like knowledge even when the answer is only a plausible pattern.
Ask together: What would count as real evidence for this answer?02Confident does not mean correct.A polished answer can still be wrong.
AI is designed to produce useful-sounding responses. It may answer confidently when information is incomplete, unclear, or invented.
Ask together: How could we check this somewhere else?03It is built to make you feel good.Helpful often means agreeable.
Tell an AI your idea is brilliant and it may praise it. Ask for flaws and it will find them. Its tone is shaped by your prompt, so its approval is not independent judgment.
Ask together: What happens if we ask it to disagree?04Assume prompts are not private.Pause before you paste.
Depending on the tool and settings, prompts, photos, files, and feedback may be stored or reviewed. Children should never share identifying or sensitive information.
Ask together: Would I be comfortable if another person saw this?05You understand things AI never will.Experience, empathy, and judgment still belong to people.
AI can assist with ideas, but it does not have relationships, lived experience, responsibility, or a real understanding of consequences.
Ask together: Who should be involved when this decision truly matters?