Preparing for the Digital Battlefield of 2026: Ghost Identities, Poisoned Accounts, & AI Agent Havoc - The Hacker News

The article highlights that agentic AI will become a significant attack vector by exploiting the "confused deputy problem," where AI agents with legitimate privileges are manipulated into performing unauthorized actions such as data exfiltration or privilege escalation. These threats, combined with large-scale account poisoning leveraging identity verification weaknesses and the persistence of "ghost identities" from past breaches in IAM systems, underscore identity as the primary point of failure in future cybersecurity.