Source: www.theverge.com/tech/1037902/google-gemini-macos-ai-computer-use
Google says Gemini can now work across a user's local files, desktop apps and the web to finish multi-step tasks. It suggests uses such as speeding up workflows, filling in forms, making presentations and listing items for sale.
Users choose which of their apps Gemini can access. While it works, Gemini runs in its own background window that shows a live preview in the chat, so people can follow each step.
Google timed the launch ahead of Apple's planned restrictions on apps that request full disk access on Mac, which Apple is tying to risks from AI agents with computer use. A preview of the related settings appeared in this week's macOS 27.3 beta.
Why it matters if you build agents
Developers building agents for macOS now have a consumer example of an agent that operates local apps and files under a permission model the user controls. Apps that depend on full disk access may face new limits from Apple, so agent builders should check how their tools request file access before those changes arrive.
Get stories like this every weekday
One short email with the day's agent tooling news.