'GhostJacking' Exposes Identity Governance Gaps in AI Agents

Summary

Security researchers have discovered a new attack technique called GhostJacking that leverages identity governance weaknesses to manipulate and hijack AI agents. The attack uses security alerts and blocked events as vectors to compromise agent behavior, exposing critical gaps in how AI systems handle identity validation and access control. This research highlights the emerging threat landscape around AI agent security and the need for stronger governance frameworks.

Entities