1 min read
|
Saved February 14, 2026
|
Copied!
Do you care about this?
Google is developing a feature called "screen automation" for its Gemini AI, allowing it to control certain Android apps for tasks like placing orders or booking rides. Users will need to supervise Gemini closely, as it can make mistakes and should not handle sensitive information. The beta also hints at a "Likeness" feature for creating 3D avatars for Google Meet.
If you do, here's more
Google is moving forward with a feature called Gemini, which will enable automation on Android devices. Details emerged from the beta version of the Google app, specifically version 17.4. The feature is codenamed “bonobo” and is designed to assist users in tasks like ordering food or booking rides through specific applications via screen automation. The groundwork for this capability is laid in Android version 16 QPR3.
While this feature offers convenience, Google emphasizes the importance of oversight. Users must supervise Gemini since it can make mistakes. There’s a safety feature allowing users to stop the automation at any point and take manual control. Privacy concerns are also addressed; if the Keep Activity setting is enabled, screenshots from apps that Gemini interacts with could be reviewed by trained personnel to enhance Google’s services. Users are advised against sharing sensitive information, such as login credentials or payment details, within Gemini chats.
In addition to the Gemini feature, the beta contains references to another integration called “wasabi,” related to 3D avatars used in Google Meet. Users may have access to prompts for this feature, which suggests a push towards more personalized interactions within Google’s ecosystem. Overall, these developments indicate Google’s focus on enhancing user experience through automation and personalization, while also highlighting the need for careful management of privacy and security.
Questions about this article
No questions yet.