Hate Making Phone Calls? Google’s AI Will Make Them for You

Google launched a new AI-powered feature in Search on Wednesday that can call local businesses, check prices and availability, and report back—all without the user ever having to make a phone call.

“Search now has the agentic capability to call local businesses using AI to check on prices and availability, saving you the hassle of tracking down information yourself,” VP of Google Search Robby Stein wrote on X. “This is rolling out in the U.S., with increased access for AI Pro and AI Ultra subscribers. Makes it much easier to cross “dry cleaning” off my to-do list.”

The new AI call feature builds on Google’s Duplex, a conversational AI launched in 2018 that makes phone calls on a user’s behalf. It uses speech recognition and context to talk naturally with businesses, then summarizes the results using AI.

“We may share your first and last name with the business being contacted, and we let users know this on the confirmation screen before they submit a request,” a Google spokesperson told Decrypt. “This helps businesses recognize that real potential customers are behind these calls and makes it easier for them to serve those users.”

In addition to the business-calling feature, Google is rolling out a new Deep Search tool powered by its Gemini 2.5 Pro model.

The features are part of Google’s shift toward artificial intelligence and AI agents designed not just to answer questions but to act on the user’s behalf.

In May, during its annual Google I/O event, Google unveiled a range of new AI features, including Veo 3, Imagen 4, and Gemini Diffusion, for its flagship AI product, Gemini. That same month, Google launched the AI Edge Gallery, a platform for distributing on-device AI models to smartphones.

However, as AI becomes increasingly capable, experts warn about the amount of data it requires to function and the hidden costs associated with agentic AI taking on more personal tasks.

“My concern is whether we’re applying the right level of oversight to the higher-risk use cases,” co-founder and CEO of AI risk management company, Monitaur AI, Anthony Habayeb, told Decrypt. “And with agentic AI systems that act on your behalf, the risks compound.”

He pointed to the growing use of AI tools that handle sensitive data—like calendars, contacts, and credit cards—as an example of how misuse or failure can escalate quickly when governance, especially in the event of a system oversight, is lacking.

According to Google, businesses contacted through the feature retain control over how they are reached and what information is shared, using settings in their Google Business Profile, including the option to opt out of automated calls.

While Google said businesses only received a user’s name, not their contact details, the feature still raised concerns about data sharing.

“Everyone wants efficiency—no one likes being on hold with a restaurant or searching for a barber while traveling,” Habayeb said.

“Consumers need to determine their risk appetite,” Habayeb added. “Some will accept the convenience and data trade-offs; others won’t. What matters most is transparency and letting the market decide from there.”