OpenAI gives Codex reusable cloud environments that work across devices
TechCrunch Sarah Perez ● Covered by 2 sources
OpenAI just made Codex work from the cloud on any device, not just a laptop. That means shared, reusable dev spaces, voice control, and security checks that keep running while you’re away.
Based on reporting by TechCrunch, Sarah Perez — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
OpenAI is pushing Codex out of the laptop and into something much more persistent. At Dev Day on Tuesday, the company said its software engineering agent can now use reusable cloud development environments that developers can reach from any device, including a phone or a remote setup in the cloud.
That’s a shift from the old model, where Codex could already launch cloud tasks but treated them more like one-off remote sandboxes. OpenAI says these new environments are meant to start faster and give teams a shared workspace with approved settings and permissions. In other words, Codex is getting a place to live instead of just a place to visit.
The update also touches the Codex CLI. Developers can now start and direct tasks with their voice, and a new /agents view is meant to make it easier to hand off work and keep track of multiple tasks at once. OpenAI said it also improved everyday chores like editing prompts, resuming sessions, and using worktrees, while giving the terminal a cleaner look for longer sessions.
Codex is also moving deeper into code review. In the ChatGPT desktop app, users can read summaries, inspect code changes, or ask Codex about possible issues before they post feedback on GitHub pull requests or GitLab merge requests. OpenAI said automatic code reviews can also run while users are away from their computers.
And then there’s the security push. A new Codex Security Cloud is designed to scan full GitHub repositories on demand, on a schedule, or whenever new commits land. Codex can then investigate findings, remove duplicates, and prepare fixes in the cloud even if the user is inactive and the laptop is shut. OpenAI said the package includes access to models from its Daybreak Blue cybersecurity initiative without a separate application.
The company also added API updates, including a Decisions API for real-time decision-making and an updated Agents API that supports computer use. Developers can also use Amazon’s Bedrock Managed Agents to build OpenAI agents that run entirely on AWS.
My take — AI-written commentary, not fact-checked reporting
This is the kind of move that makes Codex feel less like a toy in a tab and more like infrastructure. OpenAI clearly wants agents that keep working after the user walks away, which is sensible and also a little unsettling, because software that never blinks is great right up until it isn’t. The bigger trend here is obvious: the race is no longer just about smarter models, but about who gets to own the workflow around them.
Read more about this at: TechCrunch