Anthropic has expanded its computer use capability for Claude, bringing the feature to Windows after initially rolling it out on macOS. The update allows the AI assistant to carry out a variety of tasks directly on a user’s computer. This comes after the company announced the rollout via its official account on X, stating that the feature is now available through Claude Cowork and Claude Code Desktop. The capability is currently being offered as a research preview to subscribers of Claude Pro and Claude Max.
This computer use feature allows Claude to interact with the system in the same way that human users do. It can open apps, navigate software interfaces, write and test code, identify bugs, and even fix them with minimal input. According to Anthropic, users can initiate complex workflows with a single prompt, allowing the AI to manage multiple stages of app development and testing independently.
On Windows, the feature does not require extensive setup. After installation, Claude will be able to access system resources and execute tasks within the approved boundaries. Users can retain control by using session-based permissions, which means that the AI can only interact with applications that have been explicitly authorised during that session. Additional safeguards include sensitive access alerts, a single session control lock, and the ability to halt operations immediately using simple commands such as the Esc key.
The AI giant has also integrated the feature with the Dispatch tool, which allows users to assign tasks to Claude remotely via smartphone, expanding its utility beyond the desktop environment.
In separate news, the company addressed recent user concerns about Claude’s faster-than-usual usage limits. Lydia Hallie stated that users were not overcharged, citing tighter limits during peak usage periods and longer context sessions. She also recommended that Pro users choose the Claude Sonnet 4.6 model for more efficient use, noting that higher-end models tend to consume resources faster.