Palo Alto Networks security-intel boss calls AI agents 2026's biggest insider threat

An interview with Palo Alto Networks’ Wendi Whitmore argues that autonomous AI agents are becoming a new insider-threat vector as enterprises rush to adopt them, creating risks from over-privileged “superuser” agents, prompt-injection, and tool misuse. The discussion describes how attackers increasingly abuse AI to scale operations and query internal LLMs (e.g., activity documented by Anthropic), and urges CISOs to enforce least-privilege for AI identities, strong access controls, and rapid detection/response for rogue agent behavior.