Okta is spending about $200 million to bolt AI-agent security onto its identity platform, snapping up startup Permiso.
The identity giant is acquiring Permiso, a startup focused on identity threat detection, according to a source who spoke with TechCrunch. Terms weren't officially disclosed, but the source pegs the price at roughly $200 million. The deal folds Permiso's detection tools into Okta's stack, aimed at a problem enterprises are only starting to grapple with: securing non-human identities, including the AI agents now running through cloud environments. Okta hasn't said how quickly Permiso's technology will be integrated into its existing products.
AI agents multiply fast and often carry broad permissions with none of the oversight applied to human logins, making them an obvious target and an easy blind spot. For Okta, buying a threat-detection specialist is a faster route to relevance in that fight than building the capability in-house. It also signals a shift in identity security, from tracking who logged in to tracking what that identity is allowed to do and whether it's behaving normally, a much harder problem.
Expect more of these tuck-in deals as identity vendors race to cover an attack surface that barely existed two years ago.