The FBI warns of highly targeted OAuth consent phishing attacks aimed at prominent figures, their families, and acquaintances. Attackers impersonate government officials, media personnel, or known contacts, luring victims with fake file shares or meeting invitations. Granting permission to a malicious OAuth application gives attackers long-term access to emails, files, and other sensitive account data. […]
AI agents challenge traditional IAM by acting at machine speed and scale, with unpredictable access patterns that human-centric security models weren’t designed for. Prompt injection, credential exposure and multi-agent delegation create attack paths that conventional authentication and authorization controls struggle to govern. Existing behavioral monitoring can be effectively blind to autonomous agents, while ownership, accountability, […]