The companies did not disclose the details of the deal, but TechCrunch reported that it values Permiso at $200 million and is expected to close in the third quarter of Okta’s fiscal 2027.
In a blog post, Okta cited data which shows that about 58% of executives have dealt with an AI-related security incident in the past year. Okta said Permiso will bring its large research base to the deal, drawing on thousands of identity-related detection signals across dozens of partner integrations to flag issues like excessive permissions, unused access, and unusual behaviour from AI agents.
Once its capabilities are folded into Okta’s platform, they will give security teams a unified view of identity risk from human employees, service accounts, and autonomous AI agents, rather than requiring separate tools for each category. This deal also extends Okta’s reach beyond core identity management into security operations.
Okta said this will give broader visibility into how AI agents, running locally, in the cloud, or inside SaaS tools, behave. It will also incorporate a sandboxing system called SandyClaw, designed to test AI agent skills and prompts for supply-chain style attacks before they reach production.
Okta chief product officer Ely Kahn said the acquisition is central to helping organisations secure what he called the “agentic enterprise,” where humans, applications, and AI agents increasingly work side by side.
“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,” Jason Martin, cofounder of Permiso, said in the post.
The startup was founded by Martin and Paul Nguyen in 2020 and emerged from stealth mode two years later. Permiso has raised about $29 million to date, according to the report.