Military units assume attack is inevitable and rehearse constantly. Most enterprise security teams hold an annual tabletop and call it preparation. Writing in CSO Online, SimSpace co-founder and CTO Lee Rossey, previously a group leader at MIT Lincoln Laboratory, argues that treating security as a compliance checkbox rather than an operational capability is why so many teams would fail a military readiness test. The evidence keeps accumulating: ransomware disruption at retailers, insurance brokers, Jaguar Land Rover and Asahi; Cisco research showing frontier models behave significantly worse under multi-turn attack; Google’s discovery of the first zero-day exploit created using AI; Volt Typhoon inside Massachusetts utilities for months. His answer is AI Proving Grounds, validating operators and AI agents before deployment.
