Cogensec Security Research Lab
We are an enterprise AI security research lab focused on the hard problems of autonomous systems: agent integrity, adversarial coherence, multi-agent collusion, and the assurance evidence regulated deployments require.
Open research
We publish reproducible papers, open datasets, and reference implementations so the wider community can verify and build on our work.
Collaboration
We partner with frontier labs, enterprises, and academic groups on joint research, red-team exercises, and benchmark design.
Global, distributed
Our researchers and engineers work across time zones, with secure infrastructure for remote collaboration and classified-equivalent isolation when needed.
Work with us
Whether you are a researcher, a red teamer, or an enterprise team with a hard agentic-security problem, we want to hear from you.
Current focus areas
- →Agent integrity measurement and scoring
- →Sociolinguistic attacks and semantic inversion
- →Multi-agent collusion and emergent behavior
- →Constitutional AI for cybersecurity
Open positions
We are always looking for researchers, engineers, and red teamers who care about making autonomous AI trustworthy.
View careers