Gemini is more embedded in Chrome

Google is putting Gemini deeper into Chrome, turning the browser itself into the place where page summaries, tab-by-tab questions and web tasks now happen. (support.google.com) Google’s Chrome help pages say Gemini in Chrome uses the content of the current tab by default, can share up to 10 open tabs on desktop, and is separate from the standalone Gemini app. The feature is currently limited to signed-in users age 18 or older on Mac, Windows or Chromebook Plus devices in the United States, Canada, India and New Zealand, and it does not work in Incognito mode. (support.google.com) Google said in a September 18, 2025 post that Gemini in Chrome was rolling out to Mac and Windows desktop users in the United States, and a newer Google post says the experience now centers on a side panel that stays available while people move across tabs. Google says users can ask it to summarize articles, compare information across pages and handle multi-step web actions. (blog.google 1) (blog.google 2) That shifts Chrome further from being a page viewer toward being an assistant layered over the page itself. Google’s own description says the system can use page context, connect to apps such as Gmail, Calendar, YouTube, Maps, Google Shopping and Google Flights, and keep helping without forcing users into a new tab. (blog.google) (support.google.com) Google is also building administrative controls around that integration. Chrome Enterprise documentation says organizations can allow Gemini integrations, allow the feature without improving Google’s artificial intelligence models, or block the feature entirely through policy settings in the Google Admin console. (support.google.com) The privacy tradeoff is explicit in Google’s support language: Gemini in Chrome uses the current tab for responses, and desktop users can add more tabs for extra context. Google also says tabs opened from a Gemini chat or from a shared tab are shared with Gemini as well. (support.google.com) Google’s latest Chrome post also points to a more active assistant. It says “auto browse” can carry out multi-step tasks on the web, while pausing for confirmation on actions such as purchases or social posts. (blog.google) At the same time, some users are looking for ways to strip the feature back out. How2Shout said it tested Chrome 146.0.6478.127 on Windows 11 in April 2026 and found no single master switch, instead recommending separate steps for the toolbar button, Chrome’s artificial intelligence settings, the address-bar “AI Mode” flag and page-content sharing. (how2shout.com) Reports about reusable Gemini “Skills” suggest Google is also trying to make repeated browser tasks easier by letting people save prompts for later use, but that feature appears to be emerging in Chrome Canary reporting rather than in Google’s public help documentation so far. The result is a Chrome that increasingly treats browsing, asking and acting as one workflow. (ouranimeworld.com) (support.google.com)