Gemini Task Automation Launches in Beta on Galaxy S26 Ultra

Google's promised task automation for Gemini is now live in beta on Samsung's Galaxy S26 Ultra, marking the first real-world deployment of the AI assistant's ability to autonomously control apps on behalf of users.

The feature, announced weeks prior by Google and Samsung, enables Gemini to complete tasks across food delivery and rideshare applications through simple natural language prompts. Testing conducted by The Verge confirms the system works as intended, successfully handling requests like booking Ubers and ordering from Starbucks without explicit step-by-step instructions.

How It Works in Practice

When given a task, Gemini operates the target app in a virtual window, navigating menus, making reasonable decisions, and requesting clarification when needed. In testing, a request to "order an Uber to the airport" prompted the assistant to ask which airport, then automatically filled in the destination and skipped irrelevant fields like airline selection.

More complex requests requiring real decision-making also performed well. When ordering a coffee and croissant, Gemini correctly identified a flat white from Starbucks' menu and independently decided to warm the chocolate croissant—a detail the tester didn't specify but would have chosen anyway.

The system consistently halts before executing final transactions, requiring explicit user approval before confirming any order or booking. Users can watch Gemini work through each step or take manual control at any point if the automation deviates from their intent.

Current Limitations

The beta rollout is limited to food delivery and rideshare applications on the Galaxy S26 Ultra. Google has not disclosed timelines for expanding to additional apps or device models, nor specified which specific services are supported beyond generic category mentions.

Pricing for the task automation feature remains unconfirmed. It's unclear whether this will be bundled with existing Gemini subscriptions or require separate payment.

What This Means

This represents the first production deployment of agent-based UI automation on a consumer device—a capability that researchers and companies have pursued for years. The ability to autonomously navigate interfaces and make context-aware decisions demonstrates meaningful progress in AI reasoning for real-world tasks. However, the beta's limited scope and device exclusivity suggest Google is proceeding cautiously, likely gathering data on failure modes and user preferences before broader rollout. Expansion to Android devices beyond Samsung flagships and integration with more app categories will be key indicators of whether this becomes a mainstream feature or remains a premium novelty.