Okta has announced its agreement to acquire Permiso Security, an AI identity security startup, in a deal reportedly valued at just under $200 million. This strategic move underscores Okta's commitment to expanding its identity management capabilities beyond traditional user authentication to address the rapidly growing need for securing AI agents and other machine identities. As enterprises increasingly deploy autonomous software, the demand for continuous monitoring of what these non-human entities do post-access becomes paramount, intensifying competition in the identity security landscape.
Permiso's technology is designed to detect suspicious activities within cloud environments after initial access has been granted to users or applications. The startup excels at identifying attacks that leverage stolen or compromised identities to navigate cloud infrastructure. Crucially, Permiso has evolved its platform to specifically monitor AI agents and machine identities. Its innovative SandyClaw platform further enhances this by analyzing AI agent skills in a sandboxed environment, allowing for the identification of potentially malicious behaviors before these agents are deployed into live operations.
This acquisition is highly significant for developers and researchers, highlighting the critical shift in identity security towards protecting non-human entities in an AI-first world. It emphasizes that robust security frameworks must now encompass continuous threat detection and response for autonomous software, not just human users. For those building and deploying AI, it signals the necessity of integrating advanced identity threat detection and prevention capabilities from the ground up, ensuring the integrity and trustworthiness of AI systems as they become more deeply embedded across enterprise operations.