Business Security Weekly (Audio) Automating Compliance and Risk with Agentic AI as CISOs (R)Evolve - Trevor Horwitz - BSW #417
15 snips
Oct 15, 2025 Trevor Horwitz, Founder and CISO at TrustNet, brings over 20 years of cybersecurity experience to the table. He delves into how agentic AI can revolutionize compliance and risk management, automating tedious tasks and enhancing efficiency. Trevor highlights the importance of human oversight despite AI’s capabilities, discussing the fine line between automation and human involvement. He also addresses the evolving role of CISOs in navigating organizational challenges and stresses the need for authentic leadership without succumbing to bureaucracy.
AI Snips
Chapters
Transcript
Episode notes
Trevor's Security Leadership Background
- Trevor introduced himself as a CISO and founder with decades of security experience and leadership roles.
- He brings practical experience from TrustNet, iTrust, and InfraGard Atlanta to the AI and compliance discussion.
LLMs Produce Useful Draft Requirements
- Matt tested LLMs to generate requirements for AI projects and found outputs largely accurate but in need of review.
- He uses the generated content as a starting point, tweaks it, then finalizes with human oversight.
People‑Pleasing Drives Hallucinations
- AI models tend to 'people please,' which can drive hallucinations that invent supporting facts.
- That orientation makes validating AI outputs critical, especially for legal or audit content.