Google's Gemini Can Now Make Full Phone Calls to Businesses on Your Behalf via Pixel 11

Google is rolling out a new Gemini feature that lets the AI assistant place and manage entire phone calls to local businesses on behalf of Pixel 11 users, according to a report from The Verge published September 24, 2026.

The feature, described by Google as an "early experiment," allows users to delegate tasks like making a reservation, checking product stock, or rescheduling an appointment to Gemini without dialing the call themselves. Users simply tell Gemini to make the call using the Gemini app on their Pixel device.

How it works

According to Google, Gemini will call the business, introduce itself, navigate automated phone menus, wait on hold, and handle the conversation with whoever answers. Users retain a live transcript view of the call and can take over the conversation at any point if needed.

This builds on a "Call for me" feature Google introduced last year, which had Gemini call local businesses to check product availability. That earlier version did not include an option for users to take over the call mid-conversation—a capability now added in this new iteration.

Limited rollout

The feature is not broadly available. It's launching today exclusively for Gemini paid subscribers enrolled in the Phone by Google Public Beta on Pixel 11 devices in the US. Google has not disclosed a timeline for wider availability or whether the feature will expand to other Pixel models or non-subscriber tiers.

Competitive context

Google isn't alone in pursuing automated phone-call AI agents. Meta is reportedly developing a similar feature for its Muse AI agent that would let the assistant make phone calls on a user's behalf. However, according to reports cited by The Verge, Meta's system might route some of those calls through human operators behind the scenes rather than handling them autonomously end-to-end—a distinction Google has not indicated applies to Gemini's implementation.

Neither company has published benchmark data on call success rates, task completion accuracy, or how the systems handle edge cases like complex negotiations, multi-step transactions, or calls requiring payment information.

What this means

This is a feature update layered on top of Gemini, not a new model release—Google has not indicated a new underlying model powers this capability, and pricing beyond the existing Gemini subscription has not been disclosed. The practical value hinges on reliability: a voice agent that mishandles a reservation or misunderstands a menu option creates more friction than it removes, and Google's decision to keep humans able to take over mid-call signals the company isn't yet confident enough to make this fully autonomous. The restricted rollout—one phone model, one country, paid subscribers only, beta enrollment required—suggests Google is treating this as a controlled test rather than a finished consumer product. The real test will be whether businesses on the receiving end can reliably distinguish and handle AI-initiated calls, and whether call volume from bots becomes a nuisance that prompts pushback from the business side of these interactions.