Child Safety Policy
Last updated: June 2026 · Version 2026-06-v1
DRAFT — pending review by children's-privacy counsel before launch.
Safety is the design, not a setting
OpenKids AI is built for children and teens, and child safety is the first principle every feature is measured against. Children cannot browse the open internet through our coaches, cannot self-register under 13, and never see advertising or behavioural profiling.
How we protect every conversation
Every message a child sends and every reply a coach gives passes through layered safety review before it is shown. A deterministic crisis protocol responds to signs of distress with fixed, human-reviewed guidance and signposts real help.
Photos a child uploads to ask a question are used only to answer that question and are never stored afterwards.
What we never do
We never sell or share children's personal data. We never use a child's data to train advertising models. We never present AI as a human, and coaches are always clearly labelled as AI.
Reporting a concern
If you believe a child is at risk or you have a safety concern about the product, contact us at support@openkids.ai and we will respond as a priority. In an emergency, always contact your local emergency services first.