Exploit Development, Agentic Vulnerability Discovery, and Fake PoCs

Spencer McIntyre of VulnCheck joins Daniel Whitenack and Chris Benson to discuss AI-assisted vulnerability research and exploit development. The episode examines why many proof-of-concept exploits are unreliable, how initial-access intelligence helps security teams prioritise threats, and where agents can assist research. The conversation also considers the limitations of automated exploitation, the effects of AI guardrails on legitimate security work, and implications for threat intelligence.