OpenAI agreed on Thursday to acquire Ona, a 79-person cloud development environment provider formerly known as Gitpod. The deal aims to equip OpenAI's Codex with self-hosted sandboxes-giving enterprise users direct control over where autonomous AI agents execute code.
CIOs and CISOs hesitate to deploy fully autonomous agents due to operational risks. Uncontrolled AI can delete critical files, generate massive token bills, or fall victim to malicious manipulation. Ona's technology addresses these concerns by offering secure, persistent environments for agents to operate.
Securing the enterprise environment
"Ona brings the building blocks agents need for enterprise work: trusted, customer-controlled cloud environments where work continues across devices, inside the systems where software actually lives," Ona CEO Johannes Landgraf said. OpenAI said the acquisition will expand Codex beyond a single device or active session, helping organizations deploy agents securely in production.
While Landgraf did not disclose revenue figures, he said weekly Ona agent sessions grew 13 times in production since the start of the year. These sessions run at major institutions, including the oldest bank in the US, a major European pharmaceutical company, and a large Asian sovereign wealth fund. Arnal Dayaratna, research VP for software development at IDC, estimated Ona's 2025 annual revenue at roughly $7 million. He projected 2026 revenue could reach $10 million to $15 million, suggesting a potential acquisition price between $300 million and $450 million based on a standard 30-times multiple.
Competing with Anthropic
Industry analysts view the purchase as a strategic counterweight to Anthropic. Gartner said the move gives Codex essential scaling capabilities and responds to Anthropic's plan to support self-hosted sandboxes in Claude Managed Agents by May 2026. Tom Findling, CEO of Conifers.ai, said the deal is less about taking out a small competitor and more about ensuring Codex is enterprise-ready.
"The bigger issue is making Codex work inside real enterprise environments, where security, access controls, persistent cloud workspaces, audit trails, and integration with existing developer workflows matter just as much as the model itself," Findling said. Jason Andersen, principal analyst at Moor Insights & Strategy, echoed this sentiment, saying OpenAI must double down on development use cases to maintain its position. Traditional software and cloud providers are already building coding and operations infrastructure around popular models, leaving pure model providers vulnerable.
Integrating Ona allows OpenAI to offer a workspace that runs in a company's virtual private cloud. This setup enables IT teams to apply their own governance, including log management and credential management, preventing models from executing unauthorized actions. For teams evaluating these new workflows, exploring an AI Learning Path for Software Developers can provide context on how these tools integrate into existing systems.
Why this matters for IT and development professionals
The acquisition signals a shift from experimental AI coding assistants to governed, production-ready infrastructure. IT leaders will soon face pressure to evaluate vendor-specific integrated stacks against flexible, vendor-agnostic alternatives. Development teams will gain access to persistent, credential-managed sandboxes, but they must also prepare to configure and audit these environments to prevent unauthorized resource consumption or data exposure. Staying current with AI for IT & Development trends will be critical as these self-hosted agent capabilities roll out.
Your membership also unlocks:
