A research institute studying AI and the systems it touches.
LuMiNx works on artificial intelligence and cybersecurity, alongside the adjacent fields that shape them — language, governance, infrastructure, and rights.
Mission
We work at the seam between AI and security. The systems that decide, authenticate, persuade, and enforce now share one substrate. Studying either in isolation has become a category error.
Our work aims at a future built together — a future that belongs to all of humanity. Humanities and technology, in parallel.
Programme
Three threads run through everything we publish and ship.
- Identity without secrets
- Replacing shared-secret authentication with possession and presence. Passkeys, attested device identity, and the social fabric around recovery.
- Edge-native communication
- Federated, end-to-end encrypted messaging that runs on globally distributed runtimes — without a control plane, without a single jurisdiction.
- Model behaviour & safety
- How language models fail under adversarial pressure, and how their outputs interact with the law, the press, and public memory.
Stance
Open by default. We publish notes, code, and findings publicly unless doing so creates more harm than disclosure. We do not run analytics on this site. We do not use cookies. We do not embed third-party scripts.
Where we ship products, the same posture applies: minimum data, maximum transparency, no dark patterns.
Contact
- contact@luminx.one
- Security
- /security
- GitHub
- github.com/luminxtech