Okta buys AI-identity firm Permiso
Okta is acquiring Permiso Security in a deal TechCrunch values at just under $200M, betting on demand to secure AI agents and other machine identities.
Okta agreed on July 30 to acquire Permiso Security, an identity-threat detection startup, in a deal TechCrunch’s sources value at just under $200 million and structured as an almost all-cash transaction; Okta didn’t dispute that figure when asked. The deal is expected to close in Okta’s fiscal Q3 2027. Permiso, co-founded by former FireEye executives Paul Nguyen and Jason Martin, builds software that spots suspicious activity after a user, application, or AI agent has already gained authorized access to cloud infrastructure, and in April launched SandyClaw, a sandbox for analyzing AI agent skills for malicious behavior before deployment. Permiso had raised about $29 million to date, including an $18.5 million Series A in April 2024 that valued it near $80 million. Okta’s chief product officer Ely Kahn said Permiso will “extend Okta’s identity security fabric with proven identity threat detection and response capabilities.”
What it means for operators
Identity vendors are moving fast from verifying logins to continuously monitoring what AI agents and other machine identities do after they’re inside your systems, and Okta buying rather than building here signals the pace of that shift. If you’re deploying autonomous agents with standing access to internal systems, this is the kind of post-authorization monitoring layer worth budgeting for regardless of which vendor you pick.