Safe by design
We didn't bolt safety onto a social app — we never built the unsafe parts. Here's exactly how it works, in plain words.
Four decisions that carry the weight
Kids don't have accounts
Accounts belong to adults. A kid exists only as a profile inside your account — no password of their own, no sign-up path a child can walk alone.
Nowhere for strangers to find a kid
No public feed, no search, no suggested friends, no comments, no chat. Two kids connect only when their parents exchange a single-use code — in person, by text, however parents already talk.
Safety checks fail closed
If a safety check can't run, the request is refused or held — never quietly allowed through. Your kid occasionally sees "try again in a minute" instead of ever seeing something unchecked.
Your decision applies instantly
Permissions are re-checked on every single request, not cached. Switch sharing off, and it's off on your kid's very next tap — not at the next login.
How Vida's AI is screened
Every AI request passes five checks before it reaches any model:
Every screened request is recorded, and you can read the log — filtered per kid — in the grown-ups area.
What you control, per kid
Everything lives behind the grown-ups area, and everything is per kid — not per account:
Sharing, the careful way
Sharing artwork is off until a parent turns it on. Friend codes are created and redeemed only by parents — single-use, tied to one kid, expiring in 24 hours. Every shared picture is screened before another family can see it; if the screen can't run, the picture is held privately until it can.
Reactions are a fixed set of stickers — no captions, no comments, no messages between kids, ever. Either family's parent can end a friendship or delete a shared picture at any time, and reporting a picture hides it from every other family immediately while a person reviews it.
We're honest about limits
A safety page that only lists strengths isn't useful to anyone assessing real risk. We maintain a full, frank safety posture — including what doesn't protect a child yet and the order we're fixing it in — and we'd rather be asked hard questions than believed about something that isn't true.
Child-safety professionals, researchers, and law enforcement can reach us directly at safety@colorvidakids.ai. If a child is in immediate danger, contact local law enforcement and the NCMEC CyberTipline first — we are not an emergency service.