Google revealed on Wednesday a suite of enhancements to its Gemini AI capabilities within the Android ecosystem. A key highlight is an innovative method for Gemini to manage complex, multi-step actions such as arranging an Uber ride or placing a food order. These automated functions are part of several Gemini advancements rolling out, which also include broader scam detection for calls and upgraded Circle to Search functionalities allowing identification of every element visible on your phone’s display.
Google clarifies that these automation features enable users to delegate their personal tasks to Gemini. Nevertheless, the range of actions Gemini can currently execute remains somewhat restricted.
The company states that this beta feature will initially be compatible with a limited selection of applications across the food, grocery, and ridesharing sectors.
Furthermore, its availability will be restricted to the Gemini app on specific devices, such as the Pixel 10, Pixel 10 Pro, and Samsung Galaxy S26 series. Initially, it will launch exclusively in the U.S. and Korea.
Given the potential for errors in AI-driven automations, Google has implemented various safeguards. Firstly, these automations require a clear command from the device owner to initiate. During execution, users can monitor their progress live and intervene to halt the task if it errs or becomes unresponsive. Google also highlights that these processes operate within a secure, virtual environment on your phone, granting access solely to a few specific apps and not to your device’s broader data.
This functionality aligns with the increasing inclination towards leveraging AI for automating personal tasks. For example, ChatGPT enables users to set up tasks executable at scheduled or specific intervals, alongside providing an agent capable of diverse computer-based operations such as calendar navigation, slideshow creation, or code execution. Concurrently, Anthropic’s Cowork extends the prowess of its Claude AI to non-coding activities, empowering non-developers to automate routine file and task management. Furthermore, the AI utility OpenClaw gained significant attention recently for its aptitude in handling common daily chores like email dispatch, calendar organization, flight check-ins, and various other functions.
A further Gemini enhancement rolling out involves the broadened availability of its Scam Detection function for phone calls, which is now reaching Samsung Galaxy S26 series devices in the U.S. (This capability is already present on Pixel phones across the U.S., Australia, Canada, India, Ireland, and the U.K.) Google is additionally employing its on-device Gemini model to identify fraudulent texts in the U.S., Canada, and the U.K., initially on Pixel 10 series devices and soon extending to the Galaxy S26 series phones.
Lastly, Google announces that its Circle to Search feature, which allows users to perform searches using gestures such as scribbling and circling, can now identify and search for all visible elements on the phone screen, rather than being limited to a single item. This enables users to look up every garment and accessory in a favored outfit, or gain further information about multiple objects and their associated themes displayed on the screen.
Google has consistently rolled out Gemini updates to its Android platform through routine operating system upgrades and specific enhancements for its flagship Google Pixel phones, delivered via its frequent “Pixel Drops” updates. In contrast, Apple has reportedly faced challenges in deploying a more extensive suite of AI features, including an AI-enhanced Siri, with its launch recently postponed once more until later in the year.