deniz.in

Markets

Weather

Loading weather

· via The Verge

Google launches computer use in Gemini's macOS app, letting the AI drive Mac apps

Google has added computer use to its Gemini desktop app on macOS, letting the AI complete multi-step tasks across native apps, local files and the web. It is rolling out in English in the US.

Google launches computer use in Gemini's macOS app, letting the AI drive Mac apps

What the feature does

Google has introduced computer use in its Gemini desktop app for macOS, announcing the feature on Thursday. According to TechCrunch, the addition lets the AI carry out tasks for the user by navigating native Mac applications and browsing websites in the background, handling multi-step jobs that span local files, desktop apps and the web.

Google pitches the capability as a way to offload tedious, repetitive work. TechCrunch lists suggested use cases such as filling out forms, repeating the same task across a batch of files, sorting files into folders based on specific criteria, and building slideshows from local files or information in online dashboards. The Verge adds that Google also points to faster workflows, making presentations and listing items for sale.

How it works

Once a user describes what they want done, Gemini works in its own background window and streams a live preview of its actions into the chat, so the user can follow along while their screen and cursor remain free for other work, TechCrunch reports.

Users also keep control over what the agent can touch. The Verge reports that people can decide which of their apps Gemini is allowed to access, while TechCrunch notes that users can pause, stop and approve Gemini's actions before they happen — an approach consistent with how other vendors have implemented computer use.

Availability

The feature is initially available in English in the United States and requires the Gemini app for Mac plus a Google AI Plus, Pro or Ultra subscription, according to TechCrunch. The company did not say when it would expand to other markets or platforms.

Google is playing catch-up

As TechCrunch observes, Google arrives late to this capability: OpenAI, Anthropic and Meta have offered agentic features in their desktop applications for some time. The launch follows other recent moves by the company. Google only this week announced the broader rollout of Gemini's "Tasks," an in-app feature that lets the AI work more like an agent by signing into websites, making online purchases or booking appointments. It has also recently started admitting early testers to a revamped "CC" agent, announced in September, which is designed to help families stay organized across email, calendars, chats and tasks.

The timing is notable for another reason. The Verge reports that Gemini's computer use arrives ahead of new Apple restrictions for Mac apps that request "full disk access," restrictions motivated by the potential risks of AI agents capable of computer use. A preview of the new settings shipped with this week's macOS 27.3 beta.

Why it matters

Computer use is the most literal form of agentic AI: rather than suggesting an answer, the model operates the machine. That raises the stakes for permission models, and Apple's move to tighten "full disk access" rules for AI agents suggests platform owners intend to mediate that access rather than leave it to app makers. Google's decision to gate the feature behind paid AI subscriptions also signals that hands-on agent capabilities are being positioned as premium products, at least for now. For developers and IT teams, an agent that can drive native desktop apps — with user-approval checkpoints — offers a preview of what everyday desktop automation may look like across the industry.

  • #google-gemini
  • #ai-agents
  • #macos
  • #computer-use
  • #google

Related posts