What Breaks When the Model Is Fooled: Securing Agentic AI and Skills
About this episode
As AI deployments transition from basic chatbot interfaces into autonomous agents capable of executing complex real-world workflows, organizations are discovering that traditional input filtering can no longer guarantee safety. This podcast explores the critical security boundaries mapped across OWASP's main AI safety directories: the GenAI LLM Top 10, the Agentic Applications Top 10, and the newly released Agentic Skills Top 10. Join us as we dissect core vulnerabilities like malicious skills, over-privileged credentials, and supply chain compromises, outlining how security teams can build robust, system-level boundaries to control the blast radius when a model is inevitably fooled.
Sponsors:
Get every episode summarized
Each time CISO Insights: Voices in Cybersecurity publishes, we email you a written briefing from the transcript — the topics, who appeared, and any specific claims, with the ad reads skipped.
Email me new episodesFree for 3 shows. No card needed.
Hosts & guests
No transcript yet
This episode has not been transcribed. Request it and it moves to the front of the queue.
More episodes
More from CISO Insights: Voices in Cybersecurity
Wear Your Firewall: Gadgets, Gear, and Hacker Culture
CISO Insights: Voices in Cybersecurity
Unlocking the Compliance Stack: AI Drafting, Premium Templates, and Do-It-Yourse...
CISO Insights: Voices in Cybersecurity
CISO DIY: Building the Sovereign AI Security Department
CISO Insights: Voices in Cybersecurity
Beyond the Checkbox: The $12 Billion Fight to Redesign the Teen Internet
CISO Insights: Voices in Cybersecurity