Gemini AI: Google’s New Feature Automates Tasks on Smartphones

Google is expanding the capabilities of its Gemini AI assistant, moving beyond a chatbot interface to directly automate multi-step tasks within a user’s smartphone. A recently released video demonstrates Gemini autonomously completing a complex request – interpreting a group chat message to determine pizza orders and then placing the order through a food delivery app.

This development signals a shift in Google’s strategy for Gemini, aiming to transform it from a question-answering AI into a proactive assistant capable of independently executing tasks. The demonstration appears to position Google ahead of Apple, which announced similar ambitions for its Siri assistant in 2024 but has yet to deliver comparable functionality.

The new feature was previewed yesterday by Sameer Samat, president of Android at Google. Samat showcased a recorded demonstration on a phone, rather than issuing a live command to Gemini and indicated that the initial functionality extends to requesting rides from taxi services. Further capabilities are planned for future releases. Currently, the feature is limited to users in the United States and South Korea, and is available on the Google Pixel 10, Pixel 10 Pro, and Samsung Galaxy S26 series, all running the Android operating system.

The core innovation lies in Gemini’s ability to interact with other applications on the user’s device. Unlike traditional voice assistants that primarily respond to commands, Gemini can now actively control apps – a capability Google emphasizes is built with security and privacy in mind. According to Google, the system operates within a secure, virtualized environment, limiting Gemini’s access to only the applications necessary to complete the requested task. This approach aims to address concerns about AI assistants having unrestricted access to sensitive user data.

The initial beta release supports a limited set of applications within the food, grocery, and rideshare categories. In the U.S., supported apps include DoorDash, Grubhub, Instacart, Lyft, McDonald’s, Starbucks, Uber, and Uber Eats. In Korea, Kaemin and Kakao T are also supported. This phased rollout allows Google to gather user feedback and refine the system before expanding its capabilities and app compatibility.

Google has implemented several safeguards to ensure user control and transparency. Automations require explicit user initiation and can be monitored in real-time through notifications. Users can view the progress of a task, intervene if necessary, or halt the automation entirely. This level of control is intended to mitigate potential errors or unintended consequences that could arise from an AI autonomously managing tasks.

The technical implementation involves Gemini running applications within a secure “virtual window” on the user’s phone. This isolation prevents the AI from accessing other data or applications on the device, addressing a key privacy concern. The system also allows Google to tightly control the APIs and permissions available to Gemini, further limiting its potential impact on the user’s system.

This development aligns with a broader trend of integrating AI into everyday tasks. While voice assistants have long promised to simplify digital interactions, they have often fallen short of delivering truly seamless automation. Gemini’s ability to handle multi-step processes – such as interpreting a request, selecting options within an app, and completing a transaction – represents a significant step forward. However, the current limitations in app support and geographic availability suggest that widespread adoption is still some time away.

The success of this feature will likely depend on Google’s ability to expand app compatibility and maintain a high level of accuracy and reliability. Users will need to trust that Gemini can correctly interpret their requests and execute tasks without errors. The company’s emphasis on user control and transparency is a positive sign, but ongoing monitoring and refinement will be crucial to building that trust.

The introduction of Gemini’s task automation capabilities also raises questions about the future of human-computer interaction. As AI assistants become more capable of handling routine tasks, users may increasingly delegate responsibilities to these systems. This could free up time and mental bandwidth, but it also raises concerns about potential job displacement and the erosion of certain skills. The long-term implications of this shift remain to be seen, but Google’s latest move suggests that the era of truly intelligent personal assistants is drawing closer.