Anthropic enables Claude to operate computers and perform tasks

Anthropic has given its AI assistant Claude the ability to directly operate a user’s computer like opening apps, browsing the web, filling spreadsheets, and completing multi-step tasks with minimal human involvement. The announcement, made on Claude’s official X page, marks a significant step in the race among AI companies to move beyond conversational chatbots toward systems that can actually do things on your behalf.

This means tasks like filling out forms, navigating websites, managing spreadsheets or working across multiple apps can now be handled by the AI itself, rather than just guided by it.

You can now enable Claude to use your computer to complete tasks.

It opens your apps, navigates your browser, fills in spreadsheets—anything you'd do sitting at your desk.

Research preview in Claude Cowork and Claude Code, macOS only. pic.twitter.com/sVymgmtEMI

— Claude (@claudeai) March 23, 2026

From assistant to operator

Until recently, most AI tools functioned as assistants with helping users write, code or plan tasks. Claude’s update moves it closer to being an operator, capable of carrying out multi-step workflows inside real software environments.

The update enables users to assign tasks remotely, including from a phone, with Claude executing them on a connected computer. Claude can open applications, browse the internet, and interact with files such as spreadsheets.

One prompt Anthropic demonstrated in a video shows a user running late for a meeting and asking Claude to export a pitch deck as a PDF file and attach it to a meeting invite, with Claude carrying out the task autonomously.

How It Actually Works

Claude’s approach combines integrations with direct interaction. It first uses connected apps like Slack, Calendar and other integrations to complete tasks where access is already available.

When a required tool isn’t connected, Claude can ask for permission to open and use the application directly on your screen, interacting with it in real time. From there, it operates the interface the way a person would by clicking, typing, and navigating between windows.

Why this is getting attention?

The update has drawn attention because it expands AI’s role from thinking to doing. Online reactions have been immediate and wide-ranging, with many users expressing anxiety about what autonomous computer-use AI means for the future of work.

A user noted. “When Claude can now open apps, browse the web, and fill forms. Sir that's literally a job description.”

Some responses leaned into humour with comments like, “Hey claude can you join the 3pm meeting and say ‘nothing from my side’ every 15 minutes while i chill at my home.” Users are writing, “Interns watching this like” and “So Claude just became my junior.”

Another user said, “we went from “AI helps you think” to “AI just does it for you” real quick this is gonna change how people work way faster than they expect.”

Claude now uses your computer while you use your phone to tell claude to use your computer.

We have fully closed the loop.

— Youssef El Manssouri (@yoemsri) March 23, 2026

Comments sections also reflect how fast the AI Agents are developing by saying how there’s a new update every day.

Every frikin day is a feature day for Claude. pic.twitter.com/NSX35QyUtR

— Shikhar (@shikhr_) March 23, 2026

The Race to Build AI Agents

The announcement positions Anthropic as a direct challenger to OpenAI in the emerging AI agent space. Earlier this year, OpenAI’s OpenClaw demonstration went viral, showing an AI system booking restaurant reservations, managing calendar conflicts, and responding to emails with minimal human oversight.

Google has been experimenting with similar technology through Project Mariner, while Microsoft has embedded agent capabilities into Copilot for enterprise customers.

The computer use feature is currently in research preview and is available exclusively to Claude Pro and Claude Max subscribers on macOS.