Google has launched a “computer use” feature for its Gemini app on macOS. The AI assistant can control native Mac apps, local files and websites to complete multi-step tasks, such as batch-editing photos, extracting data from PDFs to fill out forms, or creating presentations from multiple sources.

Gemini works in an isolated background window, showing a live preview in the chat while leaving the user’s screen and cursor free. Users can monitor, pause or cancel tasks at any time, and must approve which apps Gemini can access.

The feature is initially rolling out in the U.S. in English to subscribers of Google AI Plus, Pro and Ultra plans.