ai security researcher
Kirill Novikov.
i work on cybersecurity and ai safety, mostly ai control and the security of ai agents.
- supported bya rapid grant from bluedot impact. part of anthropic's cyber verification program
- 83 merged fixesinto 38 projects, among them apple container, google go-containerregistry, kubernetes git-sync, nvidia nemo agent toolkit, microsoft openvmm, proton go-crypto. all of it
- contributorluthien: security reviews of an open-source ai control proxy