'GhostJacking' Exposes Identity Governance Gaps in AI Agents

New research shows how attackers can use security alerts and blocked events to manipulate and hijack AI agents.

Originally reported by darkreading on August 10, 2026. Read the full story