Google unveils Gemini 3.5 Flash with frontier AI capabilities for complex agent workflows
May 19, 2026
Google is making its biggest changes to search in over 25 years. The company is adding AI agents that work around the clock to find information and introducing a completely redesigned search box powered by artificial intelligence.
The search giant announced these features at its I/O developer conference, building on the rapid success of AI Mode, which has gained over 1 billion monthly users in just one year since launch.
AI search box gets major overhaul
Google’s familiar search box is getting its biggest update since the company’s early days. The new intelligent search box expands dynamically to give users more space to describe complex questions, moving beyond simple keyword searches.
The updated interface includes AI-powered suggestions that go further than basic autocomplete. Users can now search using multiple types of content:
- Text and images
- Files and videos
- Chrome browser tabs
The new search box is rolling out globally in all countries and languages where AI Mode is available. Google is also powering the system with Gemini 3.5 Flash, its newest AI model designed for agent-based tasks and coding.
Search agents monitor the web 24/7
Google’s new information agents represent a major shift in how people interact with search. These AI agents work continuously in the background, monitoring the entire web for updates related to specific user requests.
The agents scan multiple sources including news sites, blogs, social media posts, and real-time data on finance, shopping, and sports. When they find relevant information, they send users intelligent summaries with actionable next steps.
Practical applications include:
- Apartment hunting with specific requirements
- Tracking athlete sneaker collaborations
- Monitoring industry news or competitor updates
Information agents will first launch for Google AI Pro and Ultra subscribers this summer before expanding to more users.
Booking and shopping get smarter
Google is expanding its AI booking capabilities beyond simple restaurant reservations. The new system can handle complex requests like finding a private karaoke room for six people on a Friday night that serves late-night food.
For certain services like home repair, beauty treatments, or pet care, Google will even make phone calls to businesses on behalf of users. These booking features will roll out across the United States this summer.
The company is also adding new AI shopping agents, though it plans to share more details in a separate announcement.
Custom apps built on demand
Perhaps the most ambitious new feature involves Google’s ability to create custom mini-applications on the fly. Using its Antigravity coding platform and Gemini 3.5 Flash, Search can build personalized interfaces, visual tools, and simulations tailored to specific questions.
Users might ask for help understanding astrophysics and receive interactive visualizations, or request a custom fitness tracker that incorporates local weather data and real-time reviews of nearby gyms.
These custom experiences will be available free for everyone this summer, with more advanced mini-app creation coming first to Google AI Pro and Ultra subscribers in the US.
Personal data integration expands globally
Google is expanding Personal Intelligence in AI Mode to nearly 200 countries and territories across 98 languages, with no subscription required. This feature allows AI search to access personal information from Gmail, Google Photos, and soon Google Calendar to provide more relevant results.
The company emphasizes that users maintain full control over which apps connect to Personal Intelligence, with transparency and privacy built into the system’s core design.
Why this matters for the search industry
Google’s moves come as the search landscape faces increasing pressure from AI chatbots and alternative search engines. The company’s ability to reach 1 billion AI Mode users in one year while seeing record-high query volumes suggests its AI integration strategy is working.
These agent-based features represent a shift from reactive search to proactive information gathering, potentially changing how people interact with information online. The success of these tools could influence how other tech companies approach AI integration in their own products.
