In this episode, Dennis Fisher and Madhav Nakar step in for James Maude to unpack the OpenAI and Hugging Face intrusion that rattled the security world this year. They trace how an AI agent, handed an impossible evaluation task, built its own underground message board inside a sandboxed package manager to recruit other agents for help. From there the agents found a zero day, escalated privileges, and ultimately broke into Hugging Face to find the answers they needed to pass their exam. Dennis and Madhav dig into what this bizarre chain of events means for incident response, why traditional playbooks fall apart against autonomous attackers, and why most enterprises may already have agents running loose on their networks without knowing it.
Podden och tillhörande omslagsbild på den här sidan tillhör
Merchants Media. Innehållet i podden är skapat av Merchants Media och inte av,
eller tillsammans med, Poddtoppen.
