Safe and Secure AI

Safe and Secure AI

We are a research lab focused on artificial intelligence that is safe to deploy and secure against misuse. Our work spans robustness, alignment, and cybersecurity; developing methods, tools, and empirical understanding so AI systems can be used responsibly in the real world.

Recent publications

Selected papers from the group.

View all papers

Recent projects

Open-source tools, benchmarks, and labeled links.

View all projects