The New Attack Surface: Why AI Security Isn't App Security with Dr. Adnan Masood

Why do AI systems break decades-old security rules? Dr. Adnan Masood explains what's actually different.

Chief AI Architect at UST and co-author of Responsible AI in the Enterprise, Dr. Masood breaks enterprise AI governance into five layers, from regulatory frameworks down to day-to-day controls. He also covers what makes AI a genuinely new attack surface: a recent AI-proxy supply chain compromise, multi-turn jailbreak techniques, and why a coding agent with root access might be the least skeptical developer on your team.

See Full Show Notes and Resources

(00:00) - Welcome to Cyber Sentries

(01:55) - Meet Adnan

(04:20) - Timeline

(09:01) - Governance

(16:54) - Figuring Out What to Account For

(23:29) - How They Address It

(32:06) - Pushback

(35:41) - Taking Risks

(37:20) - Wrap Up