Google isn’t giving up its search crown anytime soon. During last year’s annual developer conference, it dipped its toes into generative AI with the launch of AI Overviews. Within a year, the platform is being used by 1.5 billion people monthly across more than 200 countries, Google CEO Sundar Pichai said.
Now, the company is going all in with a new search experience called AI Mode. “It’s getting even better, with personal context, deeper research, complex analysis and visualisation, live multimodality, and new ways to shop. Now, that’s a lot, because AI Mode can do a lot,” said Google’s VP of Search, Liz Reid, onstage at I/O.
Google stated that AI Mode is now the most powerful AI search, with more advanced reasoning and multimodality, and the ability to go deeper through follow-up questions and helpful links to the web.
With this move, the tech giant is challenging ChatGPT and Perplexity AI. The latter recently rolled out Comet, its agentic web browser, as a beta version, limited to select Apple Silicon Mac users. The browser aims to reimagine web browsing through context-aware intelligence.
Interestingly, when Perplexity CEO Aravind Srinivas was recently asked if Google would double down on AI-enabled search features that may render Perplexity “obsolete”, he said, “They [Google] have had two years to kill Perplexity and haven’t.”
Srinivas added that Google is reluctant to fully embrace AI-generated search responses because they could undermine its advertising business model.
According to a Wharton report, Google’s search engine market share remains dominant, controlling around 90% of the global search market. Last year, the company generated $198 billion from search-based advertising, accounting for more than half of its total revenue of $350 billion.
Google is also adding deep research features to AI Mode through a tool called Deep Search. It uses a method called “query fan-out,” which breaks complex questions into smaller parts and searches for each one at the same time. This helps it dig deeper into web content and give more accurate, detailed answers, along with the links for further reading.
“Google’s AI-led transformation of search should be seen as a strategic adaptation to a fragmented digital knowledge landscape. The rise of AI copilots, chat-based interfaces, and proprietary knowledge agents has redefined how users formulate queries and consume answers,” Sanchit Vir Gogia, CEO of GreyHound Research, told AIM.
Building on Deep Search, the company is also integrating Project Astra’s live capabilities into Search. With a new feature called Search Live, users can have real-time, back-and-forth conversations with Search using their camera.
Moving beyond searching the internet
Pichai said that with permission, Gemini models can tap into context from across Google apps in a way that is private, transparent and entirely under user control.
One example of this feature is personalised smart replies on Gmail, where Gemini reads entire threads to draft context-aware responses. “I can be a better friend,” said Pichai, explaining how the feature helps him reply to friends’ emails by looking into his Drive, scanning past emails for reservations, finding itineraries in Google Docs, and matching his typical greetings, tone, style, and favourite word choices.
This feature will roll out in Gmail later this year.
Moreover, Google has launched a new Deep Research feature that lets users combine public information with their own files, like PDFs and images. This helps users get a complete picture by connecting their private data with what’s happening in the wider world, all in one place.
For instance, a market researcher can upload internal sales reports and compare them with public market trends. An academic can add journal articles to improve their research. Soon, Google will also let users pull in data from Google Drive and Gmail, making finding and connecting useful information even easier.
More Than Search
Google is bringing Project Mariner’s agentic capabilities into AI Mode. Users can now make event bookings without visiting individual websites through traditional search. For instance, users can book event tickets seamlessly. AI Mode will search across platforms, compare hundreds of options in real time, handle form-filling, and show results that match the user’s request.
In addition, Google is adding a Shopping feature to Search, allowing users to add products directly, without opening separate sites. The company has also expanded its virtual try-on tool, allowing shoppers to upload a full-length photo and see how clothes would look on them.
“AI partnerships in e-commerce are no longer optional—they’re architectural. Per Greyhound Digital Pulse 2025, 64% of retail and commerce technology leaders report that AI is now embedded into customer journey design, from discovery to checkout,” said Gogia.
He further added that while not every company must partner with a foundation model provider, those who bypass the AI layer entirely may struggle to scale personalisation, optimise logistics, or manage dynamic demand.
Beyond shopping, Google’s new AI Mode also helps analyse complex data and generates custom visuals based on your query. For example, one can compare home-field advantage between two baseball teams and get an interactive graph using real-time sports data. This feature is coming to sports and finance searches.
With these updates, Google is turning Search into a tool that goes far beyond finding links.
