Gemini Makes Calls Alone: Here’s How It Works

A new capability under test on the Pixel 11 allows Google’s AI assistant, Gemini, to independently place and manage phone calls. This function enables the assistant to contact companies and stores to complete simple tasks without requiring the user to participate in the conversation directly.

How autonomous calling works

The feature targets interactions that consume significant time, such as checking product availability, finding appointment slots, or modifying existing reservations. It also addresses calls that involve navigating automated menus and waiting for a human representative. Once the task is defined, Gemini dials the number, interacts with the interactive voice response system, and waits on the line until it reaches the appropriate employee. The assistant then explains the reason for the call and requests information or executes the specified task.

Users are not excluded from the process; they can monitor the live conversation on their screen and intervene at any moment if clarification or a decision is needed. A notable aspect of this trial is transparency: Gemini identifies itself as an artificial intelligence assistant calling on behalf of the phone owner from the start of the interaction, rather than posing as a real person.

Integration with existing tools

This development builds upon previous Pixel features like Hold for Me, which manages wait times during calls, and Direct My Call, which assists with navigating phone trees. By expanding Gemini’s role beyond answering questions and executing commands to completing full tasks independently, Google is shifting the assistant toward proactive agency within the device.

The functionality remains in the testing phase. Access is currently limited to Pixel 11 users in the United States who subscribe to one of the paid Gemini plans and have joined the public beta version of the Phone by Google application.