Google Unveils AI-Powered Business Calling and Gemini 2.5 Pro in Search

On July 16, 2025, Google launched a significant update to its Search platform in the United States, introducing an AI-powered business-calling feature together with the integration of Gemini 2.5 Pro into AI Mode. This announcement, shared by Sundar Pichai, CEO of Google, and elaborated upon by Robby Stein, Vice President of Product for Google Search, exemplifies Google’s forward momentum in leveraging artificial intelligence for consumer-facing experiences.

The new calling feature allows users to delegate the task of contacting local service businesses—such as dry cleaners, auto repair shops, and pet groomers—to Google’s AI agent. After filling out a brief request via Google Search, the AI, powered by an updated version of Google Duplex and Gemini, will phone the business, introduce itself transparently as an AI from Google, and clarify queries regarding prices, availability, and services. Information is compiled and delivered directly to the user. The launch covers the contiguous United States, with exceptions in Indiana, Louisiana, Minnesota, Montana, and Nebraska due to regulatory constraints.

In extensive hands-on testing, the feature proved surprisingly reliable, though minor hiccups were noted, such as occasional missed calls by some businesses. The AI's ability to navigate real-world conversations and deliver organized, actionable information with minimal user effort stands out as a meaningful productivity boost. Speaking personally, I found the system not only efficient but also attentive to context and user intent—a refreshingly helpful alternative to waiting on hold or playing phone tag.

Alongside the business-calling feature, Google has commenced the rollout of Gemini 2.5 Pro for users with Pro and Ultra subscriptions to Google’s AI Mode in Search. Gemini 2.5 Pro significantly elevates the platform’s capabilities in research and reasoning. The model is capable of breaking complex queries into sub-tasks and synthesizing results into fully cited responses, helpful for both straightforward questions and more complicated research or coding needs. On recent benchmarks, Gemini 2.5 Pro recorded a 63.8% result on SWE-Bench Verified (coding tasks) and 86.7% on the AIME 2025 mathematics assessment, outperforming key competitors such as OpenAI’s GPT-4.5 in these technical areas.

The update takes effect at a moment when consumers increasingly expect digital agents to manage not only information retrieval but also task execution. Google’s vision, as articulated by Sundar Pichai, is to place “agentic” AI in the hands of everyday users, enabling more intuitive and efficient interactions with technology.

While adoption in certain states remains pending, Google has confirmed its plans for further geographic and feature expansion. On a personal note, I believe this transition marks a new chapter for everyday users, as commonplace chores are quietly automated by capable, context-aware AIs.

The new business-calling feature and Gemini 2.5 Pro capabilities are available as of Thursday, July 17, 2025, for the majority of U.S. users, with expanded global rollout to follow pending regulatory review and public feedback.