Google announced on Thursday, November 19, 2026, that its Gemini desktop application for macOS can now act as an AI agent that directly uses the user’s Mac to navigate native apps and browse the web. The feature, called computer use, shows a live preview of the AI’s actions while allowing users to pause, stop, or approve each step, and it is initially available in English in the U.S. for subscribers of Google AI Plus, Pro, or Ultra.
The rollout puts Google behind rivals such as OpenAI, Anthropic, and Meta, forcing it to catch up in the competitive agentic AI market where desktop automation drives enterprise adoption. By enabling multistep, on‑device tasks, Google can lock users into its subscription tier and expand its AI ecosystem beyond cloud‑only interactions.