AI agents and other non-human identities are becoming a security battleground, and Okta is paying close to $200M to arm itself for the fight.
The identity giant has signed a definitive deal to buy Permiso Security, a startup focused on threat detection for machine identities, API keys and autonomous software. Neither side published the price, but TechCrunch reported the transaction is nearly all cash at just under $200M, a figure Okta did not dispute. Closing is expected in the third quarter of fiscal 2027.
Permiso’s pitch is that bots and service accounts now vastly outnumber human logins in most cloud environments, and each one needs identity controls: verification, monitoring and revocation when it misbehaves.
The purchase fits a consolidation wave in AI security as identity platforms race to absorb agent-safety specialists. Okta CEO Todd McKinnon has bet that machine identity protection becomes a core enterprise need as companies hand more work to software that acts on its own, and the deal gives Okta detection capabilities it previously lacked in that arena.
