Updated
Updated · TechCrunch · Jul 30
Okta Buys Permiso for Nearly $200 Million as AI Agent Security Demand Grows
Updated
Updated · TechCrunch · Jul 30

Okta Buys Permiso for Nearly $200 Million as AI Agent Security Demand Grows

1 articles · Updated · TechCrunch · Jul 30

Summary

  • Just under $200 million will buy Permiso Security in an almost all-cash deal, giving Okta technology to detect threats tied to AI agents and other machine identities.
  • Third-quarter fiscal 2027 is the expected close, with Okta positioning the acquisition as a move beyond login verification into continuous monitoring of what users, apps and autonomous software do after access is granted.
  • Permiso, launched from stealth in 2022 by former FireEye executives Paul Nguyen and Jason Martin, focuses on spotting attacks that exploit stolen identities inside cloud infrastructure and recently expanded into AI-agent monitoring.
  • April brought Permiso's SandyClaw launch to test AI agent skills in a sandbox, underscoring why identity vendors are racing to secure non-human identities as enterprises deploy AI more deeply across operations.
  • About $29 million had been raised by Permiso, whose April 2024 Series A valued it at roughly $80 million post-money, highlighting the premium Okta is paying to deepen its AI security push.

Insights

How will Okta's $200 million acquisition of Permiso reshape enterprise monitoring for autonomous AI agents after initial login?
Could continuous post-authentication monitoring for AI agents introduce crippling performance overhead for enterprises deploying autonomous software?
Does the rapid surge in non-human identities render traditional multi-factor authentication obsolete in modern cloud environments?