Anthropic just gave its artificial intelligence a massive upgrade. The company announced today that its popular Claude chatbot can now directly control your computer to help you finish tasks. This new capability is rolling out to the Claude Code and Claude Cowork tools, allowing the AI to open files, browse the internet, and even run software development programs on your behalf.
This is a huge step forward for AI assistants. Until now, most chatbots could only give you instructions or write text that you had to copy and paste. Now, Claude can actually do the clicking and typing for you.
Here is how it works: When you give Claude a task, the AI will first try to use special “connectors” to link up with popular services you already use, like Google Workspace or Slack. If it can’t find a direct connection to the app it needs, Claude won’t just give up. Instead, it will take control of your screen and use the computer exactly like a human would to get the job done.
Anthropic knows that giving an AI control of your computer sounds a little scary, so they built in some safety measures. The company says Claude should always ask for your permission before it takes any major actions on your machine. Even with these safeguards, Anthropic strongly recommends that users avoid letting the AI handle highly sensitive information or important passwords, at least for now, just to be safe.
At launch, this powerful new feature will only be available to people who pay for the Claude Pro and Claude Max subscription plans. Furthermore, it is currently restricted to users running Apple’s macOS. Windows users will have to wait a bit longer to hand over the mouse and keyboard to Claude.
Anthropic considers this new computer control feature a “research preview.” This means it is still in the experimental phase, and the company expects to make plenty of adjustments and improvements based on the feedback it gets from these early users.
This update also works seamlessly with another new Anthropic feature called Dispatch. Dispatch allows you to start a conversation with Claude on your phone and then seamlessly pick it up on your desktop computer. Now, you could theoretically ask Claude a question while you are on the train, and by the time you sit down at your desk, the AI could have already opened the necessary files and started working on the problem.
This new level of automation builds on the foundation Anthropic laid earlier this year. The company introduced Claude Cowork in January as a more user-friendly version of its Claude Code tool, which was designed specifically for programmers. With this latest update, Anthropic is trying to make its AI a true digital coworker for everyone, not just software engineers.
