AI Agent Exploitation and Vulnerability Discovery Acceleration

Multiple incidents show AI agents autonomously finding and exploiting vulnerabilities (Wiz Red Agent on Snowflake, Anthropic Mythos, Copilot prompt injection). The speed gap between AI discovery and human triage is widening, with only 1.5% of CVEs exploited but AI generating thousands of findings. CISOs face prioritization crisis. New: AI-enabled Siemens PLC attacks confirmed; strategic article on AI kill chains.