OpenAI Turns Codex Cloud Environments Persistent and Reusable Across Devices

OpenAI announced on Tuesday at its Dev Day that Codex, its software engineering agent, now runs in reusable cloud development environments that developers can access from any device — a computer, a phone, or the cloud.

The change matters because it turns Codex from a laptop-bound tool with isolated, throwaway sandboxes into a persistent engineering platform that teams can share. Each cloud task was previously treated as an isolated remote sandbox. Now the environments persist and stay configurable.

Reusable environments and a shared workspace

OpenAI says developers will be able to run Codex wherever they need it. The reusable development environments are designed to help tasks start more quickly and give teams a shared workspace with approved settings and permissions, according to the company.

That last detail carries weight for enterprise adoption. Approved settings and permissions give organizations a way to standardize how an autonomous agent operates across a team, addressing a common objection to letting coding agents loose on production infrastructure.

The cloud environments update was one of several Codex announcements. OpenAI also revealed a refreshed Codex CLI, a new code review experience, infrastructure hardening tools, and additional API enhancements.

Voice control and a new /agents view for the CLI

The Codex CLI update allows developers to start and direct tasks using their voice.

A new /agents view is meant to make it easier to delegate work and track multiple tasks at the same time. OpenAI also improved everyday workflows such as editing prompts, resuming sessions, and using worktrees. The CLI is getting what the company calls a "cleaner" terminal user interface designed to make longer sessions easier to read.

The moves signal where OpenAI thinks agent interfaces are heading: away from single-threaded chat sessions and toward supervising multiple parallel tasks, with voice as a first-class input for dispatching work.

Code review inside the ChatGPT desktop app

Codex can now be pulled into a new code review experience within the ChatGPT desktop app. Users can read summaries, explore code changes, or ask Codex about potential issues before sharing feedback on GitHub pull requests or GitLab merge requests.

Automatic code reviews extend this further. Codex can perform an initial review while users step away from their computers.

For teams already reviewing code on GitHub and GitLab, the feature positions Codex as a first-pass reviewer embedded in a tool many developers already have open — a direct challenge to code review startups and GitHub's own Copilot-powered review capabilities.

Codex Security Cloud and Daybreak Blue models

On the security side, OpenAI is rolling out a new toolset called Codex Security Cloud, designed to help developers secure their infrastructure.

Developers can use the tools to scan entire GitHub repositories on demand or on a regular schedule, as well as whenever new commits arrive. Codex will then investigate findings, remove duplicates, and prepare fixes in the cloud — even when the user isn't active and their laptop is closed, OpenAI said.

Notably, the toolset includes access to the models offered through OpenAI's cybersecurity initiative Daybreak Blue, without requiring developers to submit a separate application. That removes a friction point for access to OpenAI's most security-sensitive models.

The combination of scheduled scanning, autonomous investigation, and cloud-side fix preparation points toward continuous security pipelines where agents triage vulnerabilities overnight.

API updates: Decisions API with Luna and Agents API on AWS

OpenAI also announced several API updates. A new Decisions API handles real-time decision-making. The API uses Luna to answer a set of user-defined questions with predefined answers.

The company also introduced an updated version of the Agents API that now supports computer use and the ability to use Amazon's Bedrock Managed Agents to build OpenAI agents that run entirely on AWS.

The AWS integration is significant for enterprise buyers. It lets organizations build OpenAI agents inside their existing cloud environment rather than managing separate infrastructure, a concession to how large customers actually deploy production systems.

Taken together, the Dev Day updates push Codex toward always-on operation — in the cloud, on schedules, on voice command, and on AWS — with the next question being whether enterprises will hand autonomous agents the keys to their repositories and security workflows.