At Google I/O 2025, the tech giant unveiled a suite of powerful AI-driven innovations aimed at transforming user experiences across its platforms. Among the highlights is AI Mode for Search, a new feature that enables users to enter significantly longer, more detailed queries for richer search results. Google also introduced Google Beam. In a major leap for advanced reasoning, DeepMind CEO and Nobel laureate Demis Hassabis announced Gemini 2.5 Pro Deep Think, a cutting-edge research tool leveraging parallel thinking techniques, currently being tested by a select group of trusted users. Additionally, Google Meet is rolling out a live translation feature, making multilingual communication more seamless during virtual meetings. These announcements underscore Google’s continued focus on integrating artificial intelligence across its ecosystem.
While speaking onstage, Google CEO Sundar Pichai shared that the Gemini AI chatbot app now serves more than 400 million active users each month. He also noted that AI Overviews—a feature in Google Search that delivers AI-generated summaries in response to user queries—has surpassed 1.5 billion monthly users.
Google Unveils AI Mode for Search:
Google has introduced a new feature called “AI Mode” in its Search platform, set to launch for users in the United States. Described as a fully integrated AI experience, this new tab aims to transform the way people interact with Search. With AI Mode, users can input queries that are up to three times longer than traditional search entries, enabling more complex and detailed inquiries.
Google Beam:
At I/O 2025, Google unveiled a new communication tool called Google Beam. This video chat and conferencing platform leverages cameras to generate 3D avatars of participants in real time. Serving as a simplified and more affordable version of Project Starline, Beam aims to bring immersive communication to a broader audience. It relies heavily on artificial intelligence to produce lifelike, animated visuals of users, enhancing the overall realism of the experience.
Google Unveils Gemini 2.5 Pro Deep Think:
Demis Hassabis, CEO of DeepMind and Nobel laureate, has announced Gemini 2.5 Pro Deep Think — a groundbreaking tool designed to push the boundaries of reasoning and analytical research through advanced parallel processing techniques. Due to its sophistication, Google is currently limiting access to trusted testers, with no confirmed timeline for a public release.
Google Meet is Introducing a Live Translation Feature:
Google Meet is launching a new feature that offers real-time language translation, aimed at enhancing communication for a wider audience. During a recent demo, the tool successfully replicated the speaker’s tone and pacing while translating from Spanish. Beginning this week, subscribers can access the beta version, which supports live translation between English and Spanish. More language options are expected to be added soon.
Google is bringing real-time AI camera sharing to Search:
During its I/O developer conference today, Google introduced two additional methods for using its AI-driven “Live” mode, a tool that allows users to search and inquire about anything their camera captures. This functionality is being integrated into Google Search through an enhanced AI Mode and is also launching on the Gemini app for iOS. It has already been accessible on the Android version of Gemini for about a month.
Gemini Live now being expanded to all supported Android and iOS devices:
Gemini Live, a feature that Google recently introduced for Pixel devices, is now being expanded to all supported Android and iOS smartphones via the Gemini app—which currently boasts over 400 million active users each month. This tool enables users to interact with Gemini by asking questions based on both screenshots and real-time footage captured through their phone’s camera. The rollout of Gemini Live to the Gemini app on iOS and Android begins today.