SN 1098: How worried should we be? - Unpredictable Agents
Summary
Steve Gibson and Leo Laporte discuss the security risks of new AI agents, focusing on Meta’s Muse and OpenAI’s Dots. They analyze a critical zero-day vulnerability in Muse’s macOS client that allowed unauthorized access to user accounts, alongside recent rogue AI incidents linked to the testing firm Irregular. The episode also covers the "7 Deadly Sins" hacker group, an LNG ship hack, and the FBI’s response to a site breach, concluding with a broader assessment of the current state of AI security and controllability.
Listen ad-free on Castria