Claude Can Now Control Your Computer While You Step Away

Anthropic has given its AI assistant Claude the ability to physically operate your Mac computer, opening apps, clicking buttons, filling spreadsheets, and browsing the web, while you are away from your desk. The feature, called Computer Use, launched on March 23, 2026 as a research preview for Pro ($20/month) and Max ($100–200/month) subscribers. It currently works only on macOS, with Windows support expected later in 2026.

How It Works

Computer Use turns Claude from a text-based assistant into a digital worker that interacts with your screen the way a human would. When you ask Claude to do something, it first checks whether it has a direct connector, a purpose-built integration with apps like Slack, Google Calendar, or Gmail. These connectors are fast and reliable. If no connector exists for what you need, Claude falls back to screen control: it reads your screen, moves your mouse cursor, clicks buttons, and types text.

The feature pairs with Dispatch, released on March 18. Dispatch creates a persistent conversation between your phone and your Mac. You text Claude a task from your iPhone, "Export my pitch deck as a PDF and attach it to today's meeting invite", and Claude carries it out on your desktop while you commute. When you arrive at the office, the work is done.

Anthropic's demonstrations show Claude handling: exporting documents between formats, attaching files to calendar invites, pulling data from web dashboards into spreadsheets, filling out online forms, navigating file systems to find and organize documents, and running developer tools.

Safety and Limitations

Anthropic built several safeguards into the system. Claude always asks permission before accessing a new application. It scans for prompt injection attacks, a technique where malicious content on a webpage tries to hijack the AI's actions. Certain applications are restricted by default. Users can stop Claude at any point during a task.

Anthropic is transparent about the technology's immaturity: "Computer use is still early compared to Claude's ability to code or interact with text. Claude can make mistakes." Screen-based operations are also significantly slower than direct API integrations, think seconds per click rather than milliseconds.

What Is Anthropic?

Anthropic is an AI safety company founded in 2021 by former OpenAI researchers, including siblings Dario and Daniela Amodei. The company makes Claude, which competes with OpenAI's ChatGPT, Google's Gemini, and xAI's Grok. As of early 2026, Anthropic is valued at approximately $380 billion with annualized revenue exceeding $14 billion. The company recently raised $30 billion in a Series G funding round, making it one of the most valuable private companies in the world.