Google Search has reached a usage milestone of one billion monthly users for its AI Mode feature, with query volume doubling every quarter since the tool launched last year.
At I/O today, the company introduced updates designed to integrate generative AI deeper into the search process, moving beyond simple answers to tasks requiring ongoing management and custom interfaces.
Search updates
Google is deploying Gemini 3.5 Flash as the default engine for AI Mode globally. This model is selected for its performance in coding and agent workflows.
The search box itself has been redesigned. It now expands automatically to accommodate longer queries and offers suggestions that go beyond standard autocomplete. Users can upload text, images, files, videos, or Chrome tabs as inputs. The interface is rolling out in all countries and languages where AI Mode is currently available.
Follow-up questions can now be asked directly from an AI Overview. This allows a continuous conversation where context is retained and supporting links become more relevant as the discussion progresses. This feature is live on desktop and mobile worldwide.
Search agents
Google is launching information agents for users with Google AI Pro and Ultra subscriptions this summer. These tools operate in the background to monitor specific topics across blogs, news sites, social posts, and real-time data feeds.
For example, a user asking for apartment hunting assistance can list requirements and receive notifications when new listings match those criteria. Similarly, agents can alert users when a favourite professional athlete announces a shoe collaboration.
Booking capabilities are also expanding to include local experiences and services. Users can specify criteria, such as finding a private karaoke room for six people with late-night food service, and receive pricing and availability with direct booking links. For select categories including home repair, beauty, and pet care, the system can place calls on the user’s behalf. These booking features will roll out to everyone in the U.S. this summer.
Agentic coding
Google Antigravity and Gemini 3.5 Flash capabilities are now integrated into Search to build custom responses on the fly. This includes generating visual tools, simulations, and generative UI tailored to the specific question.
Users can request layouts to visualise complex topics, such as astrophysics or the mechanics of a watch. The system assembles interactive visuals, tables, and graphs in real time. These generative UI tools will be free for everyone in Search this summer.
For ongoing tasks like planning a wedding or managing a move, Search can create custom dashboards or trackers. A user asking for a health and wellness routine could receive a custom fitness tracker coded to pull live data from reviews, maps, and weather forecasts. Custom experiences built with Antigravity will be available to Google AI Pro and Ultra subscribers in the U.S. in the coming months.
Personal intelligence
Google is expanding Personal Intelligence in AI Mode to nearly 200 countries and territories across 98 languages without a subscription requirement. Users can securely connect apps such as Gmail and Google Photos, with Google Calendar support coming soon.
The feature operates on the basis that users choose when to connect apps and retain full control over their data.
What it means
The shift moves the tool from a retrieval engine to an active assistant capable of building mini-apps and managing long-term tasks. Users gain the ability to offload repetitive work to agents that monitor data and execute bookings, while developers and power users receive tools to generate custom interfaces without writing code.
