Innovation & Tech

Google Makes Gemini Tasks Free, Letting the AI Call Shops and Pay

Google made Tasks in Gemini free across six countries, with AI-placed calls, Google Pay checkout under user-set caps — and a warning that tasks can still stall short of completion.

By Olivia Hart

4 min read

Google Makes Tasks in Gemini Free and Adds Calls and Payments
Google Makes Tasks in Gemini Free and Adds Calls and PaymentsJESHOOTS-com / pixabay

What's News

  • Google made Tasks in Gemini free on November 17, 2026 in six countries: the US, Canada, Australia, India, Japan and Korea.
  • Gemini can call businesses, check out with Google Pay under a user-set spending cap, and accept tasks via forwarded email or text.
  • AI calling is unavailable in Indiana, Louisiana, Minnesota, Montana and Nebraska; Amex card support is forthcoming.
  • Google concedes tasks can fail or stall, with some websites requiring a final user click, so automation is not guaranteed.
  • Tasks run on Gemini 4 Hydrogen, introduced November 12, 2026, building on Gemini Spark from May 19, 2026.

Google opened Tasks in Gemini to hundreds of millions of users free of charge on November 17, 2026, and the agent can now call businesses and complete purchases through Google Pay — but Google itself warns that assigning work does not guarantee it will be finished.

The free tier covers users aged 18 and over in the United States, Canada, Australia, India, Japan and Korea, according to Google's product page. Pro and Ultra subscribers in selected other countries also get access. Users flip the Gemini home-screen control from Chat to Tasks; the agent then works in the background and notifies them when results are ready.

What can Gemini actually do now?

The advanced features launch first in the United States. Gemini can call businesses, use Google Password Manager for approved account access, compare products and check out with Google Pay. Users can also assign tasks by forwarding email or text messages.

In Google's own dealership example, the agent used Maps to find nearby shops, called two dealerships simultaneously, checked Calendar availability and booked a service appointment. Gemini disclosed it was an AI on those calls, and businesses can opt out. Calling is unavailable in five states: Indiana, Louisiana, Minnesota, Montana and Nebraska.

Purchases run on the Agent Payments Protocol, with digital-token support through Mastercard and Visa on supported cards. Users set a hard spending cap that Gemini is designed to respect; receipts and tracking flow into Wallet. Google says supported cards keep their rewards and purchase protections, that American Express support is coming, and that it plans to integrate the Universal Commerce Protocol for shopping from discovery to checkout.

New input channels include forwarding email to [email protected] or verifying a U.S. phone number on an eligible carrier. Google's examples include converting school emails into calendar events and adding grocery items by text or photo to an Instacart order — before handing the user back to Instacart to complete the purchase. Native connections span Gmail, Calendar, Drive, Docs, Sheets, Slides, YouTube and Maps.

The takeaway: the free rollout puts an agent that can dial phones and spend money in front of a vast audience, with the United States as the test bed.

How does Google contain the risk?

Tasks run on Google Cloud's Agent Compute Engine inside a hardware-isolated virtual machine dedicated to the user's account. The architecture separates the agent harness, which decides the next action, from the execution sandbox. Outbound traffic passes through mediated infrastructure, and persistent credentials and cloud keys stay outside the guest environment.

Prompt-injection filters scan search results and webpages, and Safe Browsing restricts which links Gemini can open. Users can add instructions mid-task. Gemini is designed to request confirmation or pre-approval before purchases and other critical steps, while routine actions can be set to always allow. Third-party connections ship off by default, and users can disconnect apps, adjust task settings or delete chat history.

The security story is layered — but Google concedes the reliability story is not airtight.

What is the catch?

Google cautions that tasks can still fail to complete or miss their objective. On some websites, Gemini gets most of the way through a task before returning the screen to the user for a final click. In plain words: automation, sometimes; hands-off, not guaranteed.

The model under the hood is Gemini 4 Hydrogen, introduced November 12, 2026 for rapid execution across agentic workflows, with Carbon and Argon positioned for deeper, long-horizon reasoning. Google says Hydrogen improves instruction following and tool selection in the Gemini app, cutting the back-and-forth needed for a reliable output.

The lineage matters. Tasks builds on Gemini Spark, the cloud-based personal agent Google introduced on May 19, 2026, running on Gemini 3.5. Spark started with trusted testers, with a U.S. Google AI Ultra beta planned next. Its design already included Workspace integration, background operation after a laptop closed or a phone locked, recurring tasks, and confirmation requests before high-stakes actions.

The free rollout hands Google a vastly wider base for its consumer agent. The documented failures and website handoffs, meanwhile, define the gap between assigning work to Gemini and having that work done independently.

Source: unite.ai

Share this article:

More from Olivia Hart

Olivia Hart

Show full bio

Staff writer covering people, launches and day-to-day practice in entrepreneurship, startups and venture funding at Business Bearings.

2,301 articles

Related articles

« Previous articleNext article »