Claude Code Desktop now allows AI agents to control macOS interfaces

Claude

Mar 28, 2026 · Updated Apr 25, 2026

Claude Code Desktop now includes a research preview of computer use for macOS Pro and Max users. This allows the agent to move beyond the terminal to interact with native apps and simulators, bridging the gap between writing code and operating the software environment.

Anthropic released a research preview of computer use for the Claude Code Desktop app on macOS. This feature enables the agent to capture screen data and simulate mouse and keyboard actions like clicking and typing. It allows Claude to operate native applications that lack a command-line interface or API.

Most coding agents are restricted to terminal commands, preventing them from verifying work in graphical environments. By controlling the desktop, the agent can now interact with the iOS simulator or hardware control panels. This shift moves AI from a text-based assistant to an environment-aware operator.

Enable this in settings by granting macOS Accessibility and Screen Recording permissions. The system uses tiered access, restricting Claude to view-only mode for browsers and click-only for IDEs. It is available for Pro and Max subscribers and supports remote task execution through the Dispatch mobile integration.

View the full update on code.claude.com

Felix Rieseberg

@felixriesebergMar 23

Today, we’re releasing a feature that allows Claude to control your computer: Mouse, keyboard, and screen, giving it the ability to use any app. I believe this is especially useful if used with Dispatch, which allows you to remotely control Claude on your computer while you’re https://t.co/tthl6vpID2

1.5k19k

View on X

Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →

See all AI news & updates from Anthropic →

Keep reading

Claude Code Desktop Adds App Preview, PR Auto-Merge, and Session Mobility

Claude Code's desktop app now lets you preview running apps, review code diffs inline, and monitor PRs with auto-fix and auto-merge without leaving the coding session. Sessions also move between CLI, desktop, web, and mobile with a single command.

OpenAIMay 13

OpenAI Refines Codex Computer Use for Background Operation Across Mac Apps

OpenAI updated Computer Use in the Codex desktop app so agents can click, type, and execute tasks across applications without taking over the user's Mac. This enables AI to work alongside humans on the same screen and cursor. Developers now handle complex multi-app workflows without workflow interruptions.

Nous ResearchJun 2

Nous Research launches Hermes Desktop to centralize autonomous agent workflows natively

Nous Research released the public preview of Hermes Desktop, a native application for macOS, Windows, and Linux designed to run its self-improving AI agent. The shift from a terminal-based interface to a dedicated desktop environment allows users to manage complex multi-step tasks with integrated file handling and remote backend connectivity.

WarpApr 4

Warp Enables Inline Code Review and Feedback for AI Agents

Warp now supports native code reviews and inline commenting for AI agents like Claude Code and Codex directly in the terminal. This creates a structured human-in-the-loop workflow that allows developers to provide granular feedback to agents without leaving their development environment.