Agents2 min read
Gemini for Mac can now control your computer and handle multi-step tasks
Google has added computer use to the Gemini app for macOS, letting the AI complete multi-step tasks across your local files, apps and the web on your behalf. The feature is available now in the US for paid Gemini subscribers.

On this page
Google has brought computer use to the Gemini app for macOS, letting an AI agent, an AI that completes a task with several steps for you rather than just giving one answer, work through tasks across your local files, desktop apps and the web while you get on with other things.
What happened
On 19 November, Google announced that the Gemini Mac app now has a computer use feature. The company describes it as a way to "get multi-step tasks done across your local files, desktop apps, and the web." You give Gemini an instruction, and it carries out the steps itself.
While Gemini works, a live preview appears in your chat window so you can follow along. Your screen and cursor stay free the whole time. You can pause, stop or approve Gemini's actions before they happen, so you stay in control.
Google suggests the feature is useful for tasks such as:
- Filling out forms
- Repeating the same action across a batch of files
- Organising files into folders based on specific criteria
- Creating slideshows from files or information in online dashboards
- Listing items for sale
You can also choose which apps on your Mac Gemini is allowed to access.
The feature is rolling out now in English for users in the US. It requires the Gemini app for Mac and a Google AI Plus, Pro or Ultra subscription. Google has not said when it will expand to other countries or languages.
What it means for your work
- Repetitive admin tasks, such as filling the same form many times or sorting files into folders, can be handed off to Gemini rather than done manually.
- You do not have to stop what you are doing while Gemini works. Your screen and cursor remain yours.
- You keep oversight. Gemini shows you what it is doing and lets you pause or stop it at any step.
- Apple is preparing new restrictions in macOS for apps that want broad access to your computer, following concerns about AI agents with computer use. A preview of those settings appeared in the macOS 27.3 beta this week, so the permissions landscape may change.
What to do now
- Check that you have a Google AI Plus, Pro or Ultra subscription. The feature is not available on free accounts.
- Make sure you have the latest version of the Gemini app for Mac installed.
- When you first use computer use, review which apps you allow Gemini to access. Start with a small set of apps you are comfortable with.
- Try it on a low-stakes repetitive task first, such as organising a folder of files, so you can see how the live preview works before using it on anything important.
- If you are outside the US or working in a language other than English, note that the feature is not available yet. Google has not announced a timeline.
The ability to watch Gemini work step by step, and stop it at any point, is the safest way to get comfortable with handing over multi-step tasks to an AI.