AI agents create new identity security challenges for enterprises

Enterprises are facing a critical shift in security as AI agents become a primary front line for identity management. At SailPoint’s Navigate event, experts highlighted that machine identities are rapidly increasing, with some estimates suggesting 109 machine identities for every human user. This surge requires remediation at machine speed, moving toward just-in-time access and real-time decision-making rather than traditional administrative oversight.

A significant challenge identified is the rise of third-party agents that arrive embedded within existing software. While companies can secure AI they explicitly choose to deploy, many agents are inherited through product updates in platforms like Slack. These agents often bypass traditional identity infrastructure because they do not authenticate through standard single sign-on processes, leaving them invisible to current security stacks. This creates a gap where autonomous actors may gain access to production infrastructure through simple chat tool permissions.