Focuses on the intersection of frontier AI agents and traditional infrastructure security, including exploit detection, system persistence, and hardware-level attributability