Short answer: no one is watching your video chat as it happens, and there's no automated system scanning your camera feed either. What actually keeps the platform safe is simpler than that, and worth explaining properly.
There's no control room, no team of moderators sitting in on random sessions, and no automated content scanner analysing your video in real time. Technically, this isn't even a policy choice — it's how the platform is built. Video and audio on Chatzyo travel directly between your device and the other person's over a peer-to-peer WebRTC connection. That stream never passes through Chatzyo's servers at all, which means there's nothing for us to watch even if we wanted to.
This is the same underlying technology used by Google Meet and similar video tools, and it's a genuine, structural limitation, not a marketing line — we couldn't review a session after the fact even if we tried, because the video was never ours to access.
Entirely off reports. When you press the Report button during a session, that's the trigger — not before, not automatically, not based on anything detected in your video. A real person on our team then reviews the report, with a target of doing so within 2 hours.
What that person is actually looking at is the report itself: the category you selected, the session identifier, and a timestamp. Not a recording, because none exists. This is also covered in our Community Guidelines and our guide to reporting someone if you want the fuller picture.
Moderation only activates when someone reports a session. There's no background scanning happening while you chat.
A report doesn't trigger an instant ban by itself. A person looks at it and decides what, if anything, needs to happen.
Text messages do pass through our server briefly to be delivered, and a simple filter checks for things like phone numbers and email addresses being shared. That's pattern-matching on text, not anyone reading your conversation — and the messages aren't kept once the session ends.
It's a fair question, and worth being honest about the actual tradeoff. Automated content detection on live video is something larger platforms with bigger engineering teams and more resources have built, and it can catch some things faster than waiting for a person to report them. Chatzyo hasn't built that, and the peer-to-peer architecture that keeps your video private by default is also what makes that kind of scanning structurally difficult to add later without changing how video is routed in the first place.
The honest tradeoff is this: your conversations stay private and out of reach of a central server, but it also means the platform depends on people reporting problems rather than catching everything automatically. We think that's the right tradeoff for what Chatzyo is, but it is a real tradeoff, not a pure win.
If something goes wrong in a session, the platform won't catch it on its own — you reporting it is what starts the process. That's worth knowing going in: press Report the moment something happens, rather than assuming a system somewhere is already handling it.
Beyond that, the encryption WebRTC uses (DTLS-SRTP, if you want the specific term) protects your connection from being intercepted by anyone outside the conversation — your internet provider, for instance, or anyone else on the network. That part of the privacy story is real and verifiable, even though it has nothing to do with whether Chatzyo itself is watching, which it isn't.