Google LLC is rolling out “computer use” capabilities for the Gemini Mac app, enabling the assistant to interact with other apps on a user’s Mac and complete multi-step tasks in the background. The feature lets users describe what they want done, while Gemini performs actions such as clicking, typing, and filling in forms, presented with a live preview in an isolated window so users can continue working.

Across outlets, the main emphasis is on what Gemini can do: working across local desktop applications and the web, automating repetitive workflows, and supporting batch-style activities such as photo editing, preparing listings, and extracting data from documents. Several sources also note that Gemini’s access is not meant to be unrestricted: users can select which apps it can use, and it can ask for approval before key actions.

Some coverage also highlights timing and broader context. Several outlets describe the rollout as gradual and tied to specific subscription tiers in the U.S., and one source frames the update as coming shortly after Apple tightened rules for automation and app permissions on macOS. Others focus on the competitive angle, comparing it to similar AI “agent” features already available from other assistants and noting the potential pressure this could place on Apple’s Siri.