Google has taken another step in integrating artificial intelligence into mobile devices with the New autonomous capabilities of Gemini for AndroidThe assistant is no longer limited to responding to voice or text commands; it is now beginning to perform complex tasks within applications, with less direct user intervention.
These features initially arrive in beta version for a specific number of mobile phones, such as the Samsung Galaxy S26 and the Google Pixel 10. Although the rollout starts in the United States and South Korea, the move points to a future in which, also in Europe and Spain, a large part of the daily tasks on the mobile phone will be delegated to AI.
Gemini will take care of ordering a taxi for you or doing your shopping for you.
The big news is that users will be able to Delegate complete processes to Gemini without monitoring each step. On the Galaxy S26, for example, you simply need to press and hold the side button to Activate the beta version of the Gemini app and give him an order like “call a taxi to work” or “make me a shopping list”.
Once the instruction is given, the artificial intelligence It works in the background within the corresponding application. (for transportation, food, or shopping) while the person continues using their phone normally, checking email, or sending messages. From the notification bar, you can check the progress, pause what the AI is doing, or cancel it if something isn't right.
Google explains that the system is capable of navigate autonomously through compatible appsSelecting products, completing forms, and leaving everything ready for the user to review and confirm at their convenience. This is a significant change from traditional assistants, which typically only opened apps or suggested actions without going that far in the execution process.
For now, this automation It is limited to certain categories of applications —primarily food, transportation, and online shopping— and only in the initial testing phase. Although the beta is launching in the United States and South Korea, it is expected that, if the system works well, Google will extend these capabilities to more regions, including the European market.
Deeper integration of Gemini within Android
With this update, Android takes a leap towards a deeper integration of Gemini into the operating system itselfIt's not just a chatbot that you ask questions to, but a component that better understands the device's context and acts based on what happens on the screen or in the apps.
Google explains that AI operates in a supervised virtual environment inside the phone, allowing the user to see in real time what Gemini is doing and, if desired, stop the taskIn this way, the company is trying to make it clear that automation does not imply losing control over what happens on the device.
This “depth” of integration seeks reduce the number of manual steps that needs to be done for the same management: fewer changes between applications, fewer repetitive forms and fewer commands that users type or dictate over and over again.
In practice, Gemini goes from being an accessory function to settling in as an active intermediary in the mobile experience, which coordinates different apps to complete tasks that previously required a lot of screen time.
Enhanced security: advanced detection of telephone scams
Another important aspect of this update is the improved protection against telephone scamsAndroid integrates advanced fraud detection features supported by Gemini, starting with the Samsung Phone app on the Galaxy S26.
When a call is received, the system is able to Analyze in real time if the content sounds suspicious. —for example, if bank details are requested or urgent payments are insisted upon— and, if it detects a possible attempt at deception, it sends an audible and haptic alert to the user.
Google emphasizes that all this analysis It is done directly on the deviceThe audio is not sent to the server, in order to preserve privacy. Furthermore, the feature is automatically deactivated when dealing with contacts saved in the address book, to avoid unnecessary surprises during regular calls.
The company has also strengthened the Detecting fraudulent messages in the Google Messages appThis expands the identification of potential phishing campaigns and suspicious communications. For now, these features are being launched in English and in the United States, but they comply with European regulatory requirements for data protection and security, so it wouldn't be surprising to see them adapted for the European market later on.
“Circle search” becomes more useful for fashion
Along with Gemini's autonomous capabilities, Android improves the visual function known as “Search with a circle” (or “Surround to Search” in some versions). This tool allows you to circle any element that appears on the screen to obtain related information, without leaving the app you are using.
With the new update, this technology will be able to recognize complete outfitsnot just individual items. That is, if the user sees a photo with a complete outfit—t-shirt, pants, sneakers—they can circle the image and the AI will suggest visual matches for each piece separately.
The system displays suggestions for similar garments, style suggestions, and purchasing options, bringing you even closer. e-commerce to direct mobile experienceIn practice, it works like a kind of integrated Google Lens, but designed to act instantly on what is already being seen on the screen.
This improvement will also be activated on mobile devices such as the Samsung Galaxy S26 and the Google Pixel 10These platforms are becoming one of the main showcases for Gemini's new features on Android. For markets like Spain, where fashion and online shopping are significant, these types of features could be especially appealing once they become more widespread.
Gemini as an increasingly autonomous assistant
The update marks a new stage for Gemini on Android, where the assistant It evolves from a simple conversational tool to a direct actor on the deviceIt no longer just tells you how to do something, but can do it itself within the apps, with final user supervision.
Google insists that the approach remains gradual: new capabilities are They are first launching as a beta on a limited number of devices and countries, with the aim of refining the system's behavior and resolving potential problems before a larger expansion.
Meanwhile, other manufacturers are also making moves in mobile artificial intelligence, which makes it clear that The competition to lead in smartphone AI is in full swing.In this context, Gemini's proposal seeks to differentiate itself through its autonomy and stronger integration with native Android functions.
Although most of these new features are currently being tested outside of Europe, their arrival foreshadows a scenario in which routine tasks such as managing purchases, booking transport, or filtering out scams They could be operated almost entirely without touching the screen, something that could significantly change the way mobile phones are used on a daily basis.
With this whole package of improvements, Android and Gemini are aiming for a mobile model in which the Artificial intelligence is taking over more and more routine workFrom the security of calls and messages to the organization of purchases and daily plans, while the user retains the final say on what is done and how it is done.
