Artificial intelligence is gradually moving beyond answering questions and generating content. Google's latest update takes this a step further: Gemini can now make certain business calls on a user's behalf.

The feature allows Gemini to contact participating businesses to find information such as prices, availability and appointment details, potentially saving users from spending time on hold or making repeated calls. The development is part of a broader shift toward AI agents that can perform tasks rather than simply respond to questions.

What can Gemini do?

Instead of manually searching for a business's phone number and calling it, users can ask Gemini to find information from a business.

For example, a user could ask Gemini to find out:

• Whether a restaurant has a particular item available

• The price of a service

• Whether a business has an appointment slot

• Other information that normally requires a phone call

Gemini can then handle the interaction with the business and return the information to the user.

Why is this different from a normal AI chatbot?

A conventional chatbot generally gives you information based on what it already knows or what it can find online.

An AI agent can go one step further by taking an action.

In this case, Gemini isn't simply telling you a business's phone number. The system can potentially make the call and interact with the business to obtain the information you requested.

This represents an important change in how people may use AI assistants in the future. Instead of asking, "How do I do this?", users may increasingly ask AI to actually do it for them.

You may not need to wait on hold

One of the biggest advantages could be saving time.

Imagine you want to know whether a salon has an appointment available tomorrow. Traditionally, you would search for the number, call the salon and wait for someone to answer.

With an AI calling feature, the assistant could potentially handle that conversation and return the result to you.

The same concept could be useful for restaurants, repair services, clinics and other businesses that rely heavily on telephone enquiries.

AI agents are becoming the next big technology trend

The development comes as the technology industry increasingly focuses on "agentic AI."

These systems are designed to perform multiple steps to accomplish a task rather than simply generate a response.

For example, an AI agent could potentially:

1. Understand what the user wants.

2. Search for the relevant business.

3. Contact the business.

4. Ask the required questions.

5. Understand the response.

6. Present the result to the user.

This could make AI assistants much more useful for everyday tasks.

But there are important limitations

AI making calls on behalf of people also raises questions about accuracy, privacy and consent.

A business employee may misunderstand an AI-generated request, while an AI system could potentially misunderstand information given during a conversation.

There is also the question of whether businesses will know that they are speaking with an AI rather than a human.

For these reasons, such features are likely to have restrictions around what kinds of calls they can make and what actions they can take.

Could AI eventually make bookings and payments?

This is where the technology could become much more powerful.

If AI agents become capable of safely handling identity verification, permissions and payments, users could eventually delegate entire tasks to them.

For example, instead of saying:

"Find me a restaurant."

A future AI assistant could potentially:

"Find a restaurant, check availability, reserve a table and add it to my calendar."

The technology industry is already moving toward this model, although sensitive actions such as payments require much stronger safeguards.

What does this mean for smartphone users?

The smartphone could gradually become less about opening individual apps and more about telling an AI assistant what you want.

Instead of switching between Google Search, Maps, phone calls, websites and booking apps, users could potentially describe the desired outcome and let an AI agent coordinate the individual steps.

Google's latest calling capability is therefore more than just another Gemini feature. It offers a glimpse of how smartphones could evolve from tools that respond to commands into assistants that actually complete tasks for users.

Bottom line: The next stage of AI may not be about generating better answers—it may be about getting things done. Google's ability to let Gemini interact with businesses by phone is an early example of that shift toward task-performing AI agents.