Key Takeaways
OpenAI will acquire Ona to enhance Codex capabilities.
Ona brings secure persistent environments to support long-running agents.
IT and engineering leaders will gain more control for secure, scalable agent workflows.
OpenAI announced on June 11, 2026, that it will acquire Ona, folding the startup's secure cloud execution and orchestration technology into its Codex ecosystem. The deal remains subject to customary closing conditions, including regulatory approvals.
According to OpenAI, more than 5 million people use Codex each week, a figure the company said is up 400% from earlier this year. Ona claims to have helped 2 million developers work in secure, reproducible cloud environments.
The acquisition aims to extend Codex beyond single-device or active-session work, letting AI agents operate persistently inside a customer's own cloud. After closing, the Ona team will join OpenAI's Codex group to advance enterprise execution capabilities.
"Enterprises want powerful agents that can do real work while meeting the security and control requirements of their environments. Ona will help us make Codex easier to deploy securely across production workflows for customers operating at the highest standards of trust and scale."
- Thibault Sottiaux
Core Products Lead, OpenAI
What the Acquisition Brings to the Table
The Ona acquisition targets several enterprise needs, according to OpenAI:
| Acquired Capability | What It Does |
|---|---|
| Persistent agent execution | Agents continue working after sessions end or devices close |
| Customer-controlled environments | Agents run inside an organization's own cloud |
| Security and governance controls | Credential scoping, activity logging and access management |
| Production workflow scaling | Moves enterprise agents from pilots to deployment |
Enterprises Still Struggle With Agentic AI
Long-running AI agents require infrastructure governance, credential controls and accountability, and most enterprises aren't ready.
The Production Gap
Only 10% of agentic AI deployments have moved beyond experimentation into production. Security, privacy and reliability remain the primary barriers, the exact domains where cloud execution architecture must deliver.
The data picture is equally stark. Only 22% of companies are prepared for AI-era demands, while 51% remain stuck with disconnected systems.
Governance as a Value Driver
The business case for tighter infrastructure control is measurable. Companies deeply committed to AI and data sovereignty reported roughly 5x higher ROI from generative AI.
Recent OpenAI News
OpenAI has executed one of the most aggressive expansion campaigns in enterprise AI history, combining acquisitions with massive infrastructure commitments.
The company acquired analytics platform Statsig for $1.1 billion in September 2025, model governance firm Neptune for under $400 million in December 2025 and red-teaming specialist Promptfoo for $86 million in March 2026.
On the infrastructure side, Nvidia committed up to $100 billion and became OpenAI's primary chip supplier, and OpenAI signed a $38 billion cloud deal with AWS to diversify beyond Microsoft. A restructured Microsoft agreement preserved a $250 billion Azure commitment while granting both parties greater independence.
On the product side, GPT-5 launched in August 2025, followed by an in-app store inside ChatGPT and GPT-5.4 with a one-million-token context window. Most recently, OpenAI launched a $4 billion enterprise deployment unit and acquired Tomoro to accelerate commercial adoption.
The capital story may be the most impressive: OpenAI closed a $122 billion private funding round in April 2026, described as the largest in history, anchored by Amazon, Nvidia and SoftBank.