OpenAI adds background agent capabilities to its Codex coding tool

On Thursday, OpenAI announced new features and updates for Codex, its automated coding tool. The update introduces background agent capabilities, allowing Codex to operate in the background on a user’s computer. These agentic features—referring to AI tools capable of performing tasks autonomously—allow the software to deploy multiple agents that work in parallel on a user’s machine without interfering with other active applications. This launch comes as OpenAI has more fiercely competed with Anthropic in recent months. Currently, Anthropic seems to be winning the war over AI-coding tools with its product Claude Code, which TechCrunch reported on last week.

The updated Codex includes several new features designed to make it a more multifaceted tool for corporate workflows:

• Background Operations: Codex can now operate in the background on a user’s computer, opening desktop applications and carrying out operations with a cursor that clicks and types.
• In-App Browser: The tool features an in-app browser that allows users to issue commands for Codex to carry out on specific web applications. OpenAI plans to eventually expand Codex’s browser capability beyond web applications on localhost—the default address for a local computer network.
• Memory: A feature that allows Codex to recall previous work sessions and generate context about how a particular user works.
• Image Generation: A new capability that can be used to create product concepts, slide visuals, mockups, and placeholder images.
• Plug-in Integrations: OpenAI has announced 111 plug-in integrations from applications including CodeRabbit and GitLab Issues, allowing Codex to carry out tasks involving those tools.

Some of the powers added to Codex seem to resemble those previously released by Anthropic. Last month, Anthropic announced that its Claude and Cowork products could remotely control a user’s Mac and desktop. The new features represent a strategic shift for OpenAI as it focuses on enterprise capabilities and retreats from consumer tools like Sora 2.

While expanding its enterprise software lineup, OpenAI continues to face legal challenges. The company is currently navigating controversies, including lawsuits over ChatGPT’s alleged mental health impact on some users.

Why it matters

This update signals OpenAI’s pivot toward making Codex a multifaceted, agentic tool for corporate workflows, directly challenging Anthropic’s recent gains in the AI-coding space.