Identity giant Okta announced on July 30, 2026 that it has signed a definitive agreement to acquire Palo Alto-based Permiso Security, a cloud-native identity threat detection and response (ITDR) startup, in a deal that sources told TechCrunch is worth approximately $200 million. The deal is expected to close in the third quarter of Okta's fiscal year 2027 and folds Permiso's runtime detection stack into Okta's identity security fabric.
What Permiso brings to Okta
Permiso, co-founded in 2020 by former FireEye executives Jason Martin and Paul Nguyen, sits in the post-authentication seat of the identity stack — the phase most modern intrusions have shifted toward. Its platform draws on more than 2,500 research-driven signals across 70-plus identity partners to detect overprivileged access, unused permissions, anomalous agent behavior, policy violations, and high blast-radius activity in real time. Its P0 Labs research team published early work on LLMjacking and continues to track emerging AI attack patterns.
The acquisition also brings SandyClaw, a dynamic sandbox that detonates AI agent skills and prompts to catch supply-chain attacks before they reach production, into Okta's product line.
Post-authentication is where the fight is now
Okta's own 2026 research found that 58% of executives reported an AI-related security incident or near miss in the past twelve months, and the company has been steadily building beyond authentication into governance and threat detection. Chief Product Officer Ely Kahn said Permiso will "extend Okta's identity security fabric with proven identity threat detection and response capabilities." Enterprise customers are already validating the fit: Autodesk Chief Trust Officer Sebastian Goodwin said Permiso lets his team "discover and monitor all identities across our environment," and framed the deal as "an important part of our broader cloud security strategy."
Consolidation in the agentic security stack
The Permiso deal comes days after Onyx Security raised a $113 million Series B from Bessemer Venture Partners to build an AI-agent control plane, and only weeks after Anthropic disclosed a red-team breach of three organizations using its Claude models — evidence that the agentic security category is hardening fast around a handful of platforms. For Okta, buying Permiso doubles as a defensive move: it pushes the company deeper into the Security Operations Center, an area rivals Permiso's own research has repeatedly probed as the next enterprise chokepoint. Okta said the transaction is not expected to change guidance issued on May 27, 2026.
Reporting based on coverage from Okta, TechCrunch, and CyberScoop.