Google's Gemini AI update to handle app tasks automatically

Google is actively developing Gemini, and an upcoming app update will enable the AI to handle numerous tasks without requiring manual smartphone interaction. Previously, Gemini could only perform basic functions like text correction and solving math problems. Now, the artificial intelligence is poised to take control of installed applications.

Dubbed "bonobo," this feature was spotted in the beta version of Google 17.4 and is currently available in Google Labs' experimental mode. It will allow Gemini to automatically interact with apps on the device. For instance, you could request a ride through Uber or order food via Uber Eats without opening the apps or manually selecting parameters.

Initially, this automation will support a limited set of applications where users are most likely to utilize it. Google will likely roll out the function first in its own apps before expanding to third-party ones. For it to work, Android 16 QPR3 or higher will be required.

Users are cautioned that Gemini may make errors. Any actions performed by the agent on your behalf will be your responsibility. If something goes wrong, you can stop the execution and take over the task. Google also emphasizes privacy: it's not recommended to share login credentials or payment information with Gemini. With the "Save activity" option enabled, trained specialists may review screenshots of Gemini's actions to improve the service.

This feature resembles existing capabilities in Google Chrome, where Gemini 3 on the Pro and Ultra versions automates routine tasks like form filling. Now, similar automation will be available in the mobile version of Gemini.

With these changes, users will be able to simply give commands to Gemini to order rides, food, and handle other repetitive tasks. Manual smartphone interaction for these actions will become virtually unnecessary.