Google is turning Gemini into something closer to a personal assistant that can actually get things done. The AI assistant can now handle everyday tasks such as calling businesses to schedule appointments, analyzing utility bills, shopping for products, and organizing household errands, reducing the amount of time users spend managing routine responsibilities.
Announced on November 17, 2026, the expanded capabilities build on Google’s push toward agentic AI, where assistants do more than answer questions or generate information. Instead, users can hand over a task to Gemini and let it work through the necessary steps, with access to connected apps and services.
The new experience is powered by Google’s Gemini 4 Hydrogen model and is rolling out free of charge to hundreds of millions of users in the U.S., Japan, South Korea, Canada, Australia, and India. Google is also introducing additional advanced capabilities, initially available in the U.S.
Gemini can call businesses and schedule appointments
One of the more interesting additions is Gemini’s ability to make phone calls on behalf of users. Rather than searching for a business, dialing its number, waiting on hold, and negotiating an appointment, users can describe what they need and let Gemini handle the conversation.
Google’s Josh Woodward, vice president of Labs, Gemini app, and AI Studio, shared an example involving a vehicle that needed servicing. He asked Gemini to call two nearby dealerships and find an early Friday morning appointment within the next month.
Using information available through Google’s Personal Intelligence and Google Maps, Gemini identified the dealerships and called both simultaneously. It navigated the conversations, handled the waiting time, found a suitable appointment, checked the user’s Google Calendar for availability, and added the confirmed booking.
The entire process took less than five minutes.
Gemini also identifies itself as an AI when speaking with businesses. Companies can opt out of receiving these calls, and the assistant won’t disclose private account or payment information without permission.
However, phone calling isn’t available everywhere in the U.S. Google says the feature is currently unavailable in Indiana, Louisiana, Minnesota, Montana, and Nebraska.
It can analyze bills and help reduce household expenses
Gemini is also getting better at handling financial chores that usually involve digging through emails, PDFs, and online account portals.
For example, Woodward described how the assistant helped investigate an unexpectedly high $389 electricity bill. After receiving permission to access the utility portal through a new Google Password Manager integration, Gemini examined previous billing statements stored in Gmail.
It identified the household’s heat pump as the primary contributor to the higher bill, explaining that it was running around 4 pm, during peak electricity pricing.
The idea is to make sense of information users already have stored across Google’s services rather than requiring them to manually compare documents and billing histories.
Gemini can also help users keep track of medical bills, insurance renewals, and other household responsibilities.
Shopping and grocery errands are becoming agentic
Shopping is another area where Google wants Gemini to take on more responsibility.
The assistant can compare products using Google’s Shopping Graph, ask follow-up questions about requirements such as color and wattage, and help complete purchases through Google Pay.
Google is introducing the Agent Payments Protocol (AP2) to support these transactions. The company is working with Mastercard and Visa, with American Express support coming soon, to enable secure payments using digital tokens on supported cards.
Users can establish spending limits before shopping, and purchases can continue earning eligible card rewards, cashback, and purchase protections. Receipts and tracking information are also sent directly to Google Wallet.
Google is also integrating grocery shopping into Gemini through its connected apps. For example, users can ask the assistant to add items to an Instacart order by typing a message or even sending a photograph of an empty product box.
Gemini can then update the shopping list and hand the user over to Instacart to complete the order.
Google says Gemini now connects with more than 60 popular apps, with developers able to apply to join its Connected Apps Platform.
Gemini can organize schedules through email and messaging
The assistant is also becoming more useful for coordinating family schedules and managing information scattered across email conversations.
Users can forward emails to gemini@google.com, allowing the assistant to extract important dates and add them to Google Calendar. It can also identify scheduling conflicts and suggest possible solutions.
In Google’s example, Woodward forwarded several school-related emails to Gemini. The assistant extracted the relevant dates, added them to the calendar, invited his wife, and identified a conflict with basketball practice. It even suggested contacting another parent to arrange a carpool change.
Google is also introducing an RCS integration that allows users to communicate with Gemini through text messages on supported devices.
In the U.S., users can set up this functionality by opening Gemini on Android or iOS, navigating to Settings, selecting Channels, and choosing Set up under Text with Gemini. A phone number verification step is required, and users need an eligible mobile carrier.

Google emphasizes security and user control
Giving an AI assistant access to emails, calendars, passwords, payment methods, and other personal information naturally raises privacy and security questions.
Google says Gemini uses its Agent Compute Engine, which runs tasks inside a dedicated, hardware-isolated virtual machine assigned to an individual account. These environments are sandboxed and aren’t shared with other users.
The company also uses multiple security layers, including prompt-injection filters that scan webpages and search results, alongside Google Safe Browsing to restrict potentially harmful links.
Users can intervene while Gemini is working, provide additional instructions, and approve important actions. For example, Gemini is designed to request confirmation or pre-approval before making a purchase. For routine activities, users can enable an Always allow setting to reduce repeated permission requests.
Google also acknowledges that the technology isn’t perfect. Some websites may prevent Gemini from completing certain actions, requiring users to take over and click the final button themselves. The company says it is working with websites and security providers to improve compatibility with AI assistants.
Users can also adjust task settings, disconnect connected Google and third-party apps, and delete their chat history.
How to use Gemini’s new Task mode
Getting started is relatively straightforward. Open the Gemini app and switch the home-screen toggle from Chat to Task. Instead of simply asking a question, describe the task you want the assistant to complete.
For supported users in the U.S., there are two additional ways to hand over tasks: forwarding emails to gemini@google.com or texting Gemini through its new messaging integration.
The expanded experience marks an important shift in how Google wants people to use its AI assistant. Gemini is moving beyond conversational assistance toward handling multi-step responsibilities across apps, websites, phone calls, and payment services.
There is still a distinction between asking an AI how to do something and trusting it to do the work. Google is now trying to close that gap, one everyday errand at a time.
Discover more from GadgetBond
Subscribe to get the latest posts sent to your email.
