OpenAI is making another major move in the race to build more capable AI agents. The company announced plans to acquire Ona, a cloud execution and orchestration startup. OpenAI is expanding Codex into a more powerful platform for long-running AI work.
The acquisition is meant to provide Codex with a persistent and secure cloud environment for AI agents to continue working beyond a single laptop, browser session or short task window. As a result, this could make AI agents more practical for businesses in software development, research, automation and complex enterprise workflows.
Why OpenAI Wants Ona
Ona focuses on cloud-based development environments that allow teams to run software projects in secure, reproducible spaces. OpenAI plans to integrate this technology into the Codex ecosystem to help AI agents operate within controlled cloud environments.
This matters because many AI agents today face limitations based on where they run. If an agent depends on a local machine or an active session, it cannot easily complete longer tasks. Therefore, OpenAI wants Codex to operate more like a persistent digital worker that can continue making progress over hours or even days.
With Ona, Codex could gain the infrastructure needed to access tools, maintain context, run tests, handle workflows, and support more complex tasks. Users would not need to stay connected the entire time.
Codex Is Becoming More Than a Coding Tool
Codex began as a developer-focused AI tool, but OpenAI is positioning it as a broader agentic work platform. The company says Codex is now used by millions of people each week for tasks such as research, analysis, building, and automation.
The Ona acquisition supports that shift. Instead of only helping users write or debug code, Codex could become a system that performs longer software and knowledge-work tasks in the background. Users would then monitor progress, review results, and provide direction when needed.
For engineering teams, this might be running test suites, fixing bugs, modernizing apps, fixing vulnerabilities, and complex software workflows.
What This Means for Enterprise AI Agents
Enterprise adoption of AI agents depends on more than model intelligence. Companies also need security, governance, permissions, logging, and control over where AI systems operate.
OpenAI says Ona’s customer-controlled execution model will allow agents to run inside an organization’s own cloud environment. That could help businesses maintain stronger control over infrastructure, data, credentials, and security boundaries. At the same time, they can still use OpenAI’s models and orchestration.
This is especially important for companies in regulated industries or large organizations that cannot allow AI tools to operate without clear oversight.
The Bigger AI Agent Race
The deal is a reminder of how fast the AI industry is moving from chatbots to agents that can take action. Major AI companies are now in a race to build systems that can do more than answer questions. Therefore, they need to do real work across software, business operations, research and productivity.
Buying Ona could help OpenAI bolster Codex at a time when enterprise clients want AI tools that can run safely at scale. As a result, persistent cloud environments may become a key foundation for the next generation of AI agents.
What Happens Next
The acquisition has not officially closed yet. OpenAI said the deal is subject to customary closing conditions, including regulatory approvals. Until the deal closes, OpenAI and Ona will remain separate companies.
After closing, Ona’s team is expected to join OpenAI and work with the Codex team to build secure, persistent execution capabilities for enterprise customers.
Bottom Line OpenAI’s acquisition of Ona makes the path forward for Codex and AI agents clear. The company is improving model intelligence. It is also investing in the cloud infrastructure needed for agents to run securely, continuously and at enterprise scale.
If the acquisition goes well, it could help Codex move past coding assistance. Codex could become a more persistent AI work platform for developers, businesses and organizations deploying AI agents in real-world production environments.
