Skip to content
Security
Found a problem? Tell us.
We build software that protects children, so we take a security report as seriously as anything we do. If you have found a weakness, here is how to reach us and what we promise in return.
- How to report
- Email research@predatorhunters.co.uk with what you found and how to reproduce it. We read every report. If it is sensitive, say so in the first line and we will reply with a way to share details safely.
- Safe harbour
- If you act in good faith, stay within this page, and give us a fair chance to fix things before going public, we will not pursue or support legal action against you. Tell us who to credit and we will, once a fix is out.
- In scope
- research.predatorhunters.co.uk and the open-source engine. Things we care about most: anything that could expose a child's data, weaken the on-device boundary, or get around a filter.
- Out of scope, and a hard line
- Never test against a real child, a real device in use by a child, or anyone who has not consented. Do not run denial-of-service, send spam, or use social engineering against our team. Use your own test accounts and devices.
- What to expect
- We aim to acknowledge a report within a few days. We will tell you honestly whether it is something we can fix, when, and we will keep you in the loop. We are a small team, so please be patient with timing, not with severity.
There is a machine-readable version of this at /.well-known/security.txt.
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.