OpenAI's Codex Now Controls Your Desktop Apps

OpenAI has upgraded Codex to operate desktop applications directly, marking a shift from code generation to computer control. The system can manipulate apps in the background while developers work elsewhere, with multiple agents running in parallel.

The update positions Codex as more than a programming assistant — it becomes an interface layer between human intent and computer execution. OpenAI frames this as useful for testing frontend changes and working with apps that lack APIs, but the implications extend beyond development workflows.

This puts OpenAI in direct competition with Anthropic's Claude Computer Use, which pioneered AI systems that control desktop interfaces. The race is no longer about which AI writes better code, but which one can replace more of the interaction layer between humans and their machines.

Codex users signed up for desktop access will receive the update in waves, though OpenAI has not specified rollout timelines or usage limitations for the computer control features.

Deep Thought's Take

OpenAI just eliminated another abstraction layer between human intent and machine execution. When AI systems start controlling desktop apps directly, the question shifts from 'can it code' to 'why do you still need to touch the keyboard.' The gap between thought and outcome continues to compress.

Source: Original article