Safety is the foundation.
We didn't bolt safety on. The product was designed around it from the first table in the database.
Concrete safeguards.
Both parents must approve every friendship.
A friendship only exists after the requesting parent issues it AND the receiving parent approves.
No precise GPS for children.
We store coarse location cells (~1.2 km), never raw coordinates. Adults see neighbourhoods, not pins.
AI-moderated content.
Every text passes a safety scan before reaching anyone. Bullying language is auto-flagged for review.
Block + report, always.
Parents can block any other child profile and report any content. Both surfaces are heavily rate-limited.
No DMs between unrelated children.
Communication is constrained to templated, parent-approved interactions tied to a play request.
Audit trail on everything.
Friendships, blocks, decisions, every meaningful action is logged to an append-only audit table.
How a report becomes a decision.
Every report is reviewed by a real human, not an algorithm.
- A parent or admin files a report from inside the app.
- AI safety scan + human moderator review the content together.
- Decision is recorded with a public reason and an audit-log entry.
- Repeated violations result in account closure and (where required) law-enforcement notification.
Spotted something? Email hello@outzy.app, we read every safety-related message ourselves.