OpenAI announced on June 11 that it has agreed to acquire Ona, a startup building secure cloud execution technology for AI agents. The deal will fold Ona’s team directly into OpenAI’s Codex group, the unit behind the company’s autonomous coding platform.
Codex, which launched in May 2025, runs AI coding tasks on OpenAI’s own cloud infrastructure. Ona’s technology lets those same kinds of agents operate inside a customer’s own environment, think AWS or Google Cloud Platform, with enterprise-grade security guardrails baked in.
What Ona actually does
Ona built a platform designed for a very specific problem: how do you let AI software engineers run autonomously in isolated environments without compromising security?
The answer, according to Ona’s approach, involves Virtual Private Cloud deployments with policy enforcement and audit logging. In English: the AI agent does its work inside the customer’s own walled garden, and every action it takes gets recorded for compliance purposes.
