Codex comes to the ChatGPT mobile app

Summary

OpenAI is bringing Codex to the ChatGPT mobile app. According to Axios, the update allows users to follow coding tasks, review outputs, approve next steps, and start new work from a phone.

The move pushes Codex toward a more continuous work model: the agent can keep running on a computer, devbox, or remote environment while the user steps in from mobile when a task needs direction or approval.

In practice

For developers and technical teams, this changes the relationship with coding agents. Instead of staying at a computer while a task runs, a user can start work in a primary environment and check progress later from somewhere else.

Axios frames the launch as part of the direct competition between OpenAI and Anthropic for developers and enterprise customers. The point is not only mobile convenience. It is about making coding agents more accessible, cheaper to try, and easier to fit into daily work.

What we still don't know

There is still an important question around quality and control. Approving technical work on a small screen may increase the risk of rushed decisions, especially when commands, code changes, or real systems are involved.

It is also unclear how teams will define permissions, human review, and operating limits for agents that keep working while the user is away from the computer. The usefulness is clear, but trust will depend on execution and audit practices.

Why it matters

• Coding agents are moving beyond the desktop and into the user's day-to-day workflow.
• Mobile approval makes it more plausible for agents to work in the background.
• For companies, this is strong training material for agent delegation, review, and governance.
• The competition between OpenAI and Anthropic is accelerating the shift from coding tools to operating systems for work.