Listen on Your Favorite App
Episode description
Diana Kelley, CISO at Noma Security and former Cybersecurity CTO at Microsoft, joins Mo to work through the real mechanics of LLM risk: why the context window flattens the trust boundary between system instructions and user data, why that makes reliable internal guardrails essentially impossible, and why agentic AI is less a new threat category and more a stress test for the hygiene debt organizations never fully paid off.
Meet the guest
Diana Kelley
Diana Kelley is CISO at Noma Security. She has held senior security leadership roles at Microsoft, IBM Security, Symantec, and Protect AI. She is co-author of Practical Cybersecurity Architecture, a LinkedIn Learning instructor on AI/ML security, and a 2023 Global Cyber Security Hall of Fame inductee.
SOUNDBITES
5 Risks Lurking in Your GenAI App (And How to Catch Them)
We pulled the best bits so you don't have to. All the gems, none of the filler.
COMING UP
RAISE Summit 2026
We pulled the best bits so you don't have to. All the gems, none of the filler.
GO DEEPER
It Takes AI to Break AI: The Case for AI Red Teaming
We pulled the best bits so you don't have to. All the gems, none of the filler.
Subscribe for new episodes
What’s New from Alice
Policy Once, Enforced Everywhere: Alice WonderFence Joins Databricks Unity AI Gateway
How Alice WonderFence integrates with Databricks Unity AI Gateway, and how to enforce your own AI guardrails across every model, tool, and agent in production.
AI in Finance: From Money Laundering to Deepfakes
Dr. Janet Bastiman has been making convincing deepfakes since 2017, long before most people knew the word. Now the Chief Data Scientist at Napier AI, she joins Mo to get into why fraud is actually easier to catch than money laundering, how a deepfake already talked a finance team out of millions, and why the human analysts checking AI matter more than ever.
It Takes AI to Break AI: The Case for AI Red Teaming
As AI systems gain autonomy, organizations need security approaches built specifically for AI behavior. Learn why AI-driven red teaming is becoming a critical defense layer.
Evaluation of Instagram Teen Accounts
This report evaluates default and opt-in content protections under real-world and adversarial conditions. The study examines safeguard effectiveness, resilience against attempts to surface inappropriate content, and platform improvements made following testing.