Google Gemini Screen Automation Feature: Google is working on the Gemini’s agent-based features to perform tasks directly on Android phones, according to details spotted in a beta version of the Google app. The upcoming feature, called “screen automation,” is expected to let Gemini interact with apps on behalf of users, such as placing orders or booking rides.
As reported by 9to5Google, signs of this feature were found in version 17.4 beta of the Google app. The information emerged through an APK teardown, a method used to analyze app code for features that may arrive in future updates. While these findings do not confirm an official launch, they suggest what Google is currently testing internally.
Gemini’s ‘Bonobo’ feature: Book rides and order food for you
Add Zee News as a Preferred Source
According to 9to5Google, the beta version includes references to a feature called “Get tasks done with Gemini,” which Google has internally named “bonobo.” This feature uses screen automation to help Gemini complete tasks within supported apps on a user’s device. It could allow users to ask Gemini to book an Uber to their office or order food from Uber Eats without opening the app themselves. However, early support may be limited to Google’s first-party apps beyond food delivery and ride-hailing. (Also Read: iPhone Fold leaks: Apple’s first foldable could pack its biggest battery ever; specs, price tipped)
Google’s Screen Automation feature: How it works
The Screen Automation feature lets Gemini act as an on-screen assistant. Instead of only answering questions, Gemini can perform tasks within apps, such as selecting options, confirming actions, or filling out forms. However, Google says this feature will work with only certain apps, so not all Android apps will support it at launch.
With these new capabilities comes added responsibility. Google notes that Gemini may make mistakes and places responsibility on users for any actions it takes. Users can stop the process at any time and take manual control of the task if needed. (Also Read: Google Gemini may allow users to import chats from ChatGPT, Claude, Grok, Microsoft Copilot, and other AI services: How to import)
Google Test ‘Wasabi’ feature for 3D Avatars in Gemini
Google is also developing another feature, codenamed “wasabi,” which is linked to something called “Likeness.” This beta feature appears to involve 3D avatars, similar to those already used in Google Meet on Android XR. Text strings suggest that users may be able to access or manage their avatar through Gemini prompts. Google notes that a user’s likeness can only be used by that individual. These details indicate that Google is preparing Gemini to play a more active role on Android devices.
