A Framework for AI Threat Readiness

This report explains that frontier AI models are rapidly increasing the speed and scale of vulnerability discovery and exploit development, and recommends an AI-ready security operating model built on four pillars—eliminate exposure and continuously scan with AI, accelerate patching and zero-day response, perform deep AI-driven code analysis, and detect/respond in real time—backed by automation, ownership, and measurable metrics.