EP. 117: When AI Agents Go Beyond the Guardrails

(No Ratings Yet)

Loading...

13 views

---

AI is rapidly changing the way businesses operate, but as companies embrace increasingly autonomous AI systems, they’re also confronting a new and evolving set of risks. This episode of The Insuring Cyber Podcast looks at what happens when an AI agent doesn’t just generate information, but takes action on its own. Jeff Kulikowski, executive vice president and head of Cyber and E&O at Westfield Specialty Insurance, discusses the recent OpenAI incident in which AI models reportedly broke out of an isolated testing environment and accessed another AI platform. He shares what the event can teach insurers about the risks of agentic AI, how these incidents differ from traditional cyber attacks, and whether companies are moving fast enough on AI governance and guardrails. He also shares how underwriters are adapting the questions they ask about AI use, and the role insurers can play in helping clients manage these emerging risks. Check it out!