Controlling Gaps & Validating Attack Paths: Innovations in Security Testing and Vulnerability Research

Your security tools may each be working as designed. What happens in the gaps between them?

In this technical session, OpenAI and SpecterOps will examine how frontier AI and adversary tradecraft can be combined to investigate weaknesses across security controls, connect findings into attack paths, and validate their impact in an authorized environment.

Using a bounded assessment scenario, the speakers will walk through the path from an initial finding to technical evidence, remediation, and retesting. The session will examine where AI can accelerate selected research and validation tasks, how practitioners challenge model conclusions, and why patch status alone is not the end of the investigation.

Attendees will leave with a practical framework for asking not only whether a weakness exists, but what an attacker could accomplish with it and whether the fix actually closes the path.