AI security research
Researching frontier security capabilities, identifying emerging threats, and advancing the field through collaboration with industry.
Evaluating dangerous cyber capabilities
Assessing AI systems in high-risk environments, measuring autonomous capabilities and understanding where AI introduces new attack surfaces and vulnerabilities.
Novel risks
Identifying and analyzing emerging security threats introduced by AI systems, from adversarial vulnerabilities to autonomous exploitation capabilities.
Cyber ranges
Building comprehensive cyber ranges and realistic scenarios for researching AI applications in both offensive and defensive security operations.
Monitoring threat landscape
Tracking and synthesizing developments in AI security as they unfold, providing strategic intelligence on how threats and defenses are evolving in practice.