Systems
Research you can install.
The models only matter once they reach a real device. These are the two apps that carry them. The camera comes first, the full shield comes next.
A camera that won't take an unsafe photo.
Children get pushed into taking photos they should never take. PH Camera checks every frame on the phone before anything is saved. If a shot is unsafe it never becomes a file.
The app asks for no internet permission at all, so the phone itself guarantees that nothing it sees can leave it. Nothing is stored, nothing is logged, nothing is sent. If the safety check cannot run, the camera will not save the shot. Safe is the only default.
A shield for the whole device.
PH Bulwark sits across every app and the open web. When unsafe content shows up it takes out only that part and leaves the rest of the page working, so a child keeps the sites they actually need.
When something serious appears it sends a parent a short, redacted alert with no message contents in it. Illegal material is blocked and reported as the law requires. The shield never keeps the content behind it.
One device, covered end to end.
The camera guards what a child creates. The shield guards what reaches them. Together they cover the whole phone, and both run the same models the research builds.
Backers · partners · researchers
Help us build the AI that keeps children safer.
We are a small, self-funded team, four years into this. If you fund safety research, want to build with us, or want to put these models to work protecting children, get in touch.