Gemini Helps You Do More on Your Phone

Google has expanded its Gemini AI with new features that let users command their phones to perform simple tasks—such as ordering a ride or food—directly through voice or text.
The update is still in beta testing and works on a limited set of apps that handle food, groceries, or transportation. It is currently available on select Pixel and Samsung phones in the United States and South Korea.

How It Works

• Step-by-step instructions: Users give Gemini a clear command and list the steps.
• Real-time monitoring: While the task runs, you can see Gemini’s progress and stop it if anything seems off.
• Sandboxed execution: The AI operates within a secure sandbox, so it can only interact with the specified apps and nothing else on your device.

Context in the AI Landscape

• ChatGPT can schedule jobs, manage calendars, and create slideshows.
• Anthropic’s Cowork automates file-related tasks without coding.
• OpenClaw sends emails or books flights.

Gemini’s task automation aligns with this broader trend of using AI to handle everyday chores more efficiently.

Additional Safety Enhancements

• Scam‑Detection: A new system warns users about suspicious phone calls on Samsung phones in the U.S.
• Text‑Scam Detector: Planned for broader device support soon.
• Circle‑to‑Search: Expands the on-screen search feature to cover everything visible, allowing users to tap an outfit and discover all its components or learn more about a group of objects.

These updates showcase Google’s commitment to blending convenience with robust security as AI becomes more integrated into daily life.