Cyber Risk Management Podcast podcast

EP 220: Every AI Agent Needs Its Own Name

0:00
50:31
Spol 15 sekunder tilbage
Spol 15 sekunder frem

Your firewalls, access rules, and least privilege all assume software behaves the same way every time. AI models don't. So what still works, what breaks, and what should you build next? Our guest Dan Stocker, Principal Cybersecurity Architect at JPMorgan Chase and a Cloud Security Alliance contributor, explains why every AI agent needs its own identity, what your assistant's memory costs you in privacy, and where human judgment still matters most. Your hosts are Kip Boyle, CISO with Cyber Risk Opportunities, and Jake Bernstein, Partner with K&L Gates.

 

LinkedIn: https://www.linkedin.com/in/danielastocker/

 

"Fire Doesn't Innovate" by Kip Boyle:https://a.co/d/0bYatohy

Flere episoder fra "Cyber Risk Management Podcast"