
The New CISO · 2026-06-04 · 43 min
What happens when an AI agent inside your company starts behaving like an insider threat? In part two, Steve Moore picks the thread back up with former FBI operative Eric O'Neill to explore how agentic AI is rewriting cybersecurity, the legal traps that follow a breach, and why the modern CISO must think like a spy hunter. Eric opens with a sobering reality: ransomware victims who decline to pay are re-attacked at staggering rates. He explains why criminals treat cybercrime as a business, invest weeks in reconnaissance - mapping SharePoint, harvesting file trees, and studying access patterns - and why a botched recovery hands them the same door twice. The conversation turns to the new insider threat hiding in plain sight: rogue AI agents. Eric shares a real case in which one executive's casual query exposed the next round of layoffs and triggered coordinated lawsuits. They unpack how agents inherit excessive access, how attackers hijack them once inside, and why organizations are now building insider-threat programs to monitor AI behavior.