Skip to content
AI.info

The Pulse

Google rolls out Gemini computer use on Mac for AI subscribers

Gemini can operate Mac apps and websites in an isolated background window while users work elsewhere. The English-language rollout is limited to eligible U.S. subscribers with an Apple Silicon Mac.

Google rolls out Gemini computer use on Mac for AI subscribers

Gemini can now click through Mac apps and websites in a background window while users keep working on their own screen. Google began rolling out the computer-use feature in the Gemini app for Mac on November 19, initially in English in the United States.

The rollout requires an Apple Silicon Mac running macOS Sequoia 15.0 or later. It is available to people aged 18 or older with personal Google accounts and a Google AI Plus, Pro or Ultra subscription.

Work in a separate window

Users can describe a multi-step task involving local files, desktop applications or the web. Gemini carries out the clicking, typing, menu navigation and form entry in an isolated background window, with a live preview in the chat; the main screen and cursor remain free.

Users choose which apps Gemini can access, can watch its progress and pause it, and approve key actions before they happen. Google’s help page says the feature is available in Gemini tasks for personal accounts, rather than work or school accounts.

To start an assignment, users select Task below the chat box, mention @Computer use and describe the work. Computer use must be enabled in the app’s settings. The setting also requires Accessibility and Screen & System Audio Recording to be turned on.

Permissions and examples

The first time Gemini is asked to work with an app, it asks whether to deny, allow or always allow access. An always-allow choice remains in effect unless the app is removed from the allowed list in settings. Users can also add apps to allowed or blocked lists directly.

Google says Gemini can extract weights, dimensions and addresses from local PDFs to fill out shipping quote forms. The company also suggests gathering metrics and screenshots from live dashboards across multiple regions and using them to build a presentation in Google Slides.

Another proposed use is preparing resale listings: Gemini can analyze clothing photos in a local folder, research prices and record item details before staging listings on resale platforms. For creative work, the examples include batch color-grading photos and setting up complex 3D scenes.

Limits and backup options

Google’s help page cautions that Gemini can make mistakes or take unexpected actions. It warns that untrusted content displayed in one app can manipulate Gemini into taking actions in other permitted apps, and that access to some local apps, such as code or text editors and developer tools, can enable shell execution or broader file-system changes.

The assistant can access logged-in accounts and active sessions in native Mac apps, and may be able to read or interact with visible multi-factor authentication prompts or codes. Some actions may be irreversible because of the app, the help page says.

Users can turn on an option for on-demand file backups. Gemini will try to create a temporary backup while working but cannot always do so; backups are kept for 24 hours or until the next prompt. Content in some apps may not be backed up, depending on their settings and permissions.

Sources

  1. Original source · Blog post Google blog.google

Other pages read

  1. Google Gemini Apps Help support.google.com

Companies in this story

Related stories

Explore

More articles