Vulnerability response in the AI-discovery era

Frontier AI models are accelerating the discovery of software vulnerabilities beyond the capacity of traditional CVE-driven triage and patching; defenders must centralize discovery (including non-CVE findings), prioritize by in-use/exploitability/impact, automate remediation and ownership, and prepare runtime detection and containment for vulnerabilities that will be exploited before they can be fully patched.