Android Developers Blog

TestingCatalog News 🗞

Google has announced a new era of task automation on Android, powered by its AI assistant, Gemini. Now available on Samsung Galaxy S26 and Google Pixel 10 series, Gemini can autonomously perform tasks within selected apps by taking control of the user’s screen. Initial applications include ride-sharing and food delivery services, offering users more seamless and efficient interactions.
To support these AI-driven interactions, Google is providing developers with two new capabilities. The first is AppFunctions, a Jetpack library and platform API, which allows Android apps to present structured and self-describing functions to Gemini, enabling natural language task execution directly on the device. This functionality is comparable to Mobile Cloud Platforms (MCP), but operates locally rather than relying on cloud backends.
Additionally, a new UI automation framework is being introduced, enabling Gemini to execute multi-step processes across any app without requiring specific developer integration. These features supplement each other to make AI agents more helpful and integrated within the Android ecosystem. Although currently in their early stages, these capabilities are expected to roll out more widely with the upcoming release of Android 17.

Collections

Google has announced Gemini-powered task automation for Android, initially available on Samsung Galaxy S26 and Google Pixel 10 series. Gemini can autonomously control the screen to perform tasks within apps like ride-sharing and food delivery. Two new developer capabilities are introduced: AppFunctions, a Jetpack library enabling apps to expose structured functions to Gemini for natural language task execution on-device (similar to MCP but local), and a UI automation framework that lets Gemini execute multi-step processes across any app without specific developer integration. Broader rollout is expected with Android 17.

Google Introduces Gemini-Powered Task Automation for Android Devices