When agents do the testing and humans do the coordinating, who’s actually accountable for the findings?
That’s the question we explored in Super Cyber Friday: Hacking Pentesting in the Age of Agentic AI.
Sprocket CTO Eric Sheridan joined Will Gregorian, CISO at Galileo Medical, and CISO Series host David Spark for a thoughtful discussion on what agentic testing actually changes and what it doesn’t.
Topics covered:
- What “human oversight” really means when agents are running autonomously and when oversight becomes the bottleneck
- How to maintain accountability when a finding comes from a decision no human explicitly approved
- How to build responsible AI into an offensive security product without sacrificing speed
- What real validation of an AI pentesting agent looks like
- Whether agentic pentesting changes the conversation security has with the business or simply accelerates it
If you’re running a pentest program, evaluating AI-powered security tools, or trying to understand where human judgment still needs to be in the loop, this recording is worth your time.
