Google has begun deploying Information agents in Search AI Mode to Google AI Ultra subscribers across all languages and markets, turning the I/O-announced feature into a live paid product.
The agents work continuously in the background, scanning blogs, news sites, social posts and real-time data on finance, shopping and sports, then sending detailed updates and links when new information appears.
Google pitched the tool for persistent tracking tasks such as apartment hunting or monitoring product drops from athletes, so users no longer need to repeat searches manually.
Robby Stein said this is only the first wave, with Google planning to expand access beyond AI Ultra subscribers later this summer.
Will Google's AI agents create a user utopia by bankrupting the websites they learn from?
As AI agents handle our daily thinking, are we sacrificing our own ability to think critically?
With AI agents prone to making things up, can we trust them to act on our behalf without causing disaster?
1 Billion Users and Counting: Google's AI Agents, Gemini 3.5 Flash, and the Future of Search and Privacy
Overview
Google is transforming its AI ecosystem by introducing advanced agents like AI Mode, Gemini Intelligence, and Gemini 3.5 Flash. These agents are designed to provide personalized, proactive assistance across devices and enterprise applications, all while emphasizing user control, transparency, and security. The rollout of these features is a strategic move to redefine search and digital interaction, responding to rising user expectations and the rapid adoption of AI-powered tools. This shift impacts content publishers, businesses, and users, driving a move toward generative optimization and a more agent-driven, secure, and personalized digital experience.