Google announced Thursday that its Gemini desktop application for macOS can now perform tasks on behalf of users by navigating native Mac apps and browsing websites in the background. The feature allows the AI to complete multistep workflows involving local files, applications, and the web, such as filling out forms, organizing files into folders, or creating slideshows from online dashboard information.
When the AI performs these actions, users see a live preview of the agent working within the chat interface. The system is designed to allow users to continue working on other tasks without interference, as the cursor and screen remain free. Users maintain control through options to pause, stop, or approve Gemini's actions before they are executed.
This release follows Google's recent announcement regarding the broader rollout of Gemini's "Tasks" feature, which enables the AI to act as an agent by signing into websites, making online purchases, or booking appointments. The company also recently began admitting early testers to its revamped "CC" AI agent, a tool intended to help families organize email, calendars, and chats.
Google's entry into computer-use capabilities follows similar agentic feature releases from competitors including OpenAI, Anthropic, and Meta. The new macOS functionality is initially available in English in the United States and requires a Google AI Plus, Pro, or Ultra subscription alongside the Gemini app for Mac.
Google did not provide a timeline for when the feature might expand to other platforms or international markets.













