Okta Agrees to Acquire Permiso Security for Almost $200 Million

Okta, the identity management company, has signed a definitive agreement to acquire Permiso Security, a startup that specializes in detecting identity-based threats in cloud environments, betting that demand for protecting AI agents and other machine identities will keep growing as enterprises hand more of their operations to autonomous software. The deal, announced Thursday, is valued at just under $200 million and structured as an almost all-cash transaction, according to TechCrunch, which cited a source with knowledge of the terms. Okta has not disclosed the terms itself, but a company spokesperson did not dispute the $200 million figure when TechCrunch asked CEO Todd McKinnon for comment.

What Permiso Brings to Okta

Permiso, based in Palo Alto, California, builds software that inventories human, non-human, and AI identities across cloud, SaaS, and on-premises environments, then watches what each one actually does after it has already been granted access. In its own description of its platform, Permiso calls the underlying technology its Universal Identity Graph, saying it connects every identity to the credentials it owns, the machines it creates, the agents it runs, and every action those identities take in real time, whether that is a human logging into a cloud service or an AI agent invoking a tool. The platform is organized around three functions Permiso calls Discover, Protect, and Defend.

TechCrunch reported that Permiso was co-founded in 2022 by Paul Nguyen and Jason Martin, both former FireEye executives, and has raised about $29 million to date, including an $18.5 million Series A round in April 2024 led by Altimeter Capital that valued the company at roughly $80 million on a post-money basis. Permiso says its research arm, P0 Labs, is staffed by several former Mandiant “advanced practices” leads and has built more than 1,500 detection signals.

In April, Permiso launched SandyClaw, a tool it describes as the first dynamic sandbox for AI agent skills, the downloadable capability packages that let agents call tools and APIs. Rather than just scanning a skill’s code the way most existing tools do, SandyClaw runs the skill with a live agent inside an instrumented sandbox, records every LLM decision, network call, file access, and credential-access attempt (intercepting SSL traffic to do so), and scores the result against detection engines including Sigma, YARA, Nova, and Snort. The tool accepts skills from agent frameworks including OpenClaw, Cursor, and Codex.

Extending Okta’s “Identity Security Fabric”

In its own announcement, Okta framed the acquisition as an extension of what it calls its identity security fabric. “We’re thrilled to welcome Permiso to Okta as we help companies secure their agentic enterprises where humans, applications, service accounts, and AI agents work together,” said Ely Kahn, Okta’s chief product officer. “Permiso will extend Okta’s identity security fabric with proven identity threat detection and response capabilities, and an incredible threat research and security team that will advance Okta’s threat detection and prevention capabilities.”

Permiso’s co-founders framed the deal as a chance to reach more customers than they could on their own. “Everything we’ve built at Permiso has been driven by one goal: helping customers stay ahead of identity-based threats in an increasingly complex security landscape,” said Martin. “Joining the leading, neutral identity provider means that work now reaches far more organizations than we could have on our own.” Nguyen added: “Together, we’ll help security teams gain deeper visibility into identity risks, respond faster to threats, and strengthen their overall security posture. Combining our technology and expertise with Okta’s platform is what makes comprehensive identity threat protection possible at this scale.”

Okta said that once the deal closes, Permiso’s P0 Labs research team will combine with Okta’s own threat intelligence group to sharpen detection and hunting capabilities across the identity lifecycle, and that customers will gain visibility into the behavior of AI agents running locally, on SaaS platforms, or in cloud environments, not just a record of who was originally granted access.

Another Bolt-On in a Crowded Identity Security Market

The deal continues a pattern for Okta, which acquired privileged access management startup Axiom Security in a transaction estimated at around $100 million and announced in August 2025, according to CTech. Where Axiom extended Okta’s control over who can reach sensitive systems, Permiso is built to watch what happens after that access is granted, a distinction that matters more as companies give AI agents standing permissions to act inside their environments on their own.

Okta is not the only identity vendor making that bet. Two days earlier, on July 28, data security company Cyera agreed to acquire non-human identity specialist Oasis Security for about $1 billion in its third acquisition of the year, according to TechCrunch. Taken together, the two deals point to the same shift: as enterprises deploy AI agents and machine credentials at a scale that increasingly dwarfs their human workforce, identity vendors are racing to add the runtime detection and response capabilities needed to watch those non-human identities after login, not just gatekeep them at the door.

The Okta-Permiso transaction is expected to close in the third quarter of Okta’s 2027 fiscal year, subject to customary closing conditions. Okta said in its press release that the deal is expected to have no impact on the financial guidance it issued on May 27, 2026.