Noticia
Google’s Gemini in Chrome brings agentic browsing to Android
Google is bringing Gemini-powered assistance and auto-browse to Chrome on Android, with page summaries, app connections and confirmation safeguards for sensitive actions.
Google is bringing Gemini into Chrome on Android, turning the mobile browser into a more active assistant for research, image editing and selected everyday tasks. The company announced the rollout on May 12, presenting the feature as an expansion of Gemini’s role beyond a standalone chatbot.
The update matters because it places AI help directly inside the browser window. Instead of copying text into another app, Android users can open Gemini from Chrome’s toolbar and ask questions about the page they are viewing. Google says Gemini can summarize long articles, explain complicated passages and help users investigate information without switching between apps.
Gemini works from the page on screen
On compatible Android devices, tapping the Gemini icon in the top-right corner of Chrome opens the assistant at the bottom of the display. The panel is designed to keep the current webpage visible while the user asks questions or requests a summary. That layout is particularly useful on a phone, where moving between a browser and an AI app can interrupt the task being researched.
Google also says Gemini in Chrome can connect with selected Google apps. Examples in the announcement include adding an event to Calendar, saving recipe ingredients in Keep and finding information in Gmail. These are context-aware actions: Gemini can use the page being viewed as part of the request, rather than relying only on a separate prompt typed into a chat window.
Some of these capabilities depend on Personal Intelligence, an optional feature that can tailor responses using information such as a user’s interests, hobbies, family and pets. Google describes the connection as something users choose to enable, which is important because the usefulness of a more personal assistant also depends on how much information a user is willing to connect.
Auto-browse handles selected tasks
The bigger change is auto-browse, which Google describes as an agentic browsing feature for Android. It is designed to carry out parts of a task across the web instead of simply returning instructions. Google’s examples include finding a parking space through SpotHero using details from a ticket confirmation, or asking Chrome to update a Chewy order when a pet’s food needs to change.
These examples do not mean that every website or service is supported. Google says auto-browse is rolling out with select devices and services, so availability will depend on the user’s region, subscription and the specific task. The feature is also not intended to make purchases or publish content without oversight. Google says Chrome asks for confirmation before sensitive actions such as buying something or posting on social media.
That confirmation step is the practical boundary between assistance and delegation. Chrome can navigate and prepare an action, but the user remains responsible for checking the final details. On a small screen, that means reviewing the recipient, quantity, address, price and other important fields before approving anything. The extra pause may make the process less automatic, but it reduces the chance that a mistaken interpretation becomes an unwanted transaction.
Rollout starts with strict limits
Google’s announcement sets clear initial requirements. Gemini in Chrome is scheduled to roll out in the United States to select devices running Android 12 or later, with at least 4GB of RAM. The device language must be set to English-US. Google also says the rollout starts at the end of June, while auto-browse is available to AI Pro and Ultra subscribers in the United States on select compatible devices.
Those conditions mean the announcement is not a promise that Gemini in Chrome will appear on every Android phone. Users who do not see the Gemini icon may be outside the initial region, use a different language setting, have unsupported hardware or simply be waiting for the feature to reach their account. Google has not provided a universal availability date for all Android markets in the announcement.
Image tools and security protections
Gemini in Chrome also uses Google’s Nano Banana image technology to customize visuals found on the web or create new ones from a request. Google gives examples such as turning study material into an infographic or changing the appearance of a room in an apartment listing. As with the browsing assistant, the result should be treated as generated content rather than an objective representation of the original image.
Google says the feature includes protections against prompt-injection attacks, in which malicious instructions embedded in a webpage attempt to manipulate an AI system. The company also says sensitive actions require confirmation. Those protections are useful, but they do not remove the need to treat web pages and generated responses cautiously, especially when personal data, payments or account access are involved.
Google’s official announcement positions Gemini in Chrome as an early step toward a browser that can understand context and complete more work on a user’s behalf. For Android users, the immediate appeal is straightforward: summaries, explanations and selected actions are available without leaving the page. The longer-term question is how quickly support expands beyond the initial U.S. English rollout, and how effectively Chrome can balance convenience with clear user control.