OpenAI Is Developing a ‘Persistent’ AI Agent
OpenAI is building a persistent AI agent feature for its Codex tool that can continue working autonomously without being stopped, according to code reviewed by WIRED. The feature appears designed to let Codex operate proactively in the background until explicitly told to stop.
Why this matters: A coding agent that keeps running until you tell it to stop is a meaningful shift from a tool you use to a process you have to manage. That changes what access it needs, how long it holds it, and what it touches while you are not watching. Right now most people assume AI tools do a task and wait. An agent that keeps going on its own is something closer to an unsupervised employee with access to your code, your credentials, and your systems. The accountability question is simple: if it does something wrong while you were not looking, who is responsible?
Who should care: General readers · AI governance · Policy
This summary is AI-assisted and may contain errors. It is an original briefing to help you gauge significance quickly — not a reproduction of the source. Always read the linked original before relying on it. See our methodology.