Google has introduced a new computer use feature for its Gemini macOS application, enabling the AI to perform multi-step tasks across local files, desktop software, and web browsers. The assistant executes actions such as clicking, typing, and filling out forms in the background, while providing a live preview within the chat interface.
The feature, which launched on November 19, 2026, is currently available in English for users in the United States. Access is restricted to subscribers of Google AI Plus, Pro, or Ultra plans, excluding free-tier users from the new functionality.
To ensure security, Google implemented a granular permission system where users must grant access to specific applications individually. Additionally, users retain the ability to pause or terminate any AI-driven task in real time through the live preview window.
This update marks an expansion of the Gemini macOS app, which originally debuted in April 2026 with basic automation tools. The company previously introduced Spark and Tasks features in June 2026, positioning this latest release as a step toward full desktop-wide agent capabilities.




