OpenAI has rolled out a major update to its AI coding assistant, Codex, significantly enhancing its capabilities beyond traditional code generation. The update positions Codex as a more autonomous and intelligent development partner, capable of interacting directly with computers, remembering user preferences, and automating complex workflows. The move comes as competition intensifies in the AI coding space, with players like Anthropic gaining traction through offerings such as Claude Code.
Previously known primarily for generating code snippets and assisting with programming tasks, Codex is now being transformed into a more comprehensive AI-powered assistant. The latest update enables it to support a broader range of developer workflows, including testing, debugging, and UI adjustments.
The upgrade reflects the broader industry trend where AI tools are evolving into end-to-end development assistants. With competitors offering increasingly advanced capabilities, OpenAI is aiming to strengthen Codex’s position in this rapidly growing segment.
One of the most notable additions is Codex’s ability to control a computer in the background. The AI can now view the screen, click, and type, allowing it to interact directly with applications and perform tasks autonomously.
This feature enables Codex to work across multiple tools and environments, even in cases where APIs are not available.
OpenAI, in a blog post, said, "Multiple agents can work on your Mac in parallel, without interfering with your own work in other apps. For developers, this is helpful for iterating on frontend changes, testing apps, or working in apps that don’t expose an API."
Another powerful capability is the ability to run multiple AI agents simultaneously. These agents can perform different tasks in parallel without disrupting the user’s workflow, significantly improving productivity for developers handling complex projects.
The updated Codex desktop app now includes a built-in browser. This allows users to interact with web pages directly, such as leaving comments or instructions that guide the AI in real time.
OpenAI has introduced over 90 new plugins that integrate apps, tools, and MCP servers. These plugins enable Codex to access broader contextual information and execute actions across different platforms seamlessly.
Codex can now generate and edit images using the GPT-image-1.5 model. This addition expands its functionality beyond coding, making it useful for design-related tasks such as UI prototyping and content creation.
The integration of image capabilities reflects the growing convergence of coding and design workflows, where developers increasingly rely on AI tools to handle both technical and creative aspects.
A key improvement in the update is the introduction of memory. Codex can now retain information about user preferences, previous projects, and workflows. This allows it to deliver more personalised and context-aware assistance over time.
Codex also gains the ability to schedule tasks and resume long-running processes automatically. This feature is particularly useful for developers working on extended projects, enabling the AI to continue tasks even when the user is not actively engaged.
The new features are being rolled out to users of the Codex desktop app who are signed in with ChatGPT. This ensures a phased implementation, allowing OpenAI to refine the features based on user feedback.
Initially, the computer-use feature is available on macOS, with plans to extend support to users in the EU and UK in the near future. Additionally, personalisation features such as memory and context-aware suggestions will soon be available to Enterprise and Edu users, as well as those in regulated regions.
The update signals a broader shift in the AI industry toward autonomous agents capable of performing multi-step tasks independently. This evolution could significantly change how developers interact with software tools.
By combining computer control, memory, and automation, Codex has the potential to dramatically improve developer productivity, reducing the time required for repetitive and complex tasks.
OpenAI’s latest update to Codex marks a major leap in the evolution of AI-powered development tools. By introducing computer control, memory, and advanced integrations, Codex is no longer just a coding assistant but a comprehensive digital collaborator. As competition in the AI coding space intensifies, such innovations are likely to redefine how developers build, test, and manage applications. The future of software development is increasingly autonomous, and Codex’s new capabilities offer a glimpse into that transformation.