Overview
- Google rolled out Gemini’s new “computer use” for macOS on Thursday, November 19, 2026, letting the app click, type, navigate menus and fill forms across local files, native Mac apps and websites to complete multi-step tasks.
- The feature runs in an isolated background window with a live preview in chat, gives per-app permission controls, pauses for logins and payments, and lets users pause or stop actions and confirm key steps before they execute.
- Access is limited to English users in the United States with an Apple Silicon Mac running macOS Sequoia 15.0 or later, an 18-plus personal Google account and a Google AI Plus, Pro or Ultra subscription.
- Google’s documentation warns that broad permissions can expose active app sessions and visible multifactor authentication codes, allow the agent to interact with developer tools or the shell, and produce irreversible file changes or backups that some apps do not support.
- The release follows Apple’s macOS 27.3 developer beta changes to tighten automation permissions and transparency and comes after similar agentic desktop controls from OpenAI, Anthropic and Meta, so adoption and safety will hinge on OS safeguards and how Google narrows risky access.