Google shifts search to AI mode

Google is rebuilding Search around AI Mode, and the clearest signal is not the interface but the behavior Google says it is already seeing. New company data released this week showed U.S. users are typing longer, more conversational prompts into AI Mode, while voice and image inputs are rising and planning-oriented queries are growing fastest. At the same time, Google said it is testing Gemini-powered conversational ads and two new ad formats inside Search, tying product changes directly to monetization. ### What changed in the way people are searching? Google’s first-year AI Mode data showed that users in the United States are moving away from short keyword strings and toward fuller, back-and-forth prompts. Search Engine Journal, citing Google’s figures, reported that visual and voice-based searches are becoming more prominent and that planning and decision-making queries are changing fastest. One cited figure said planning-related queries grew 80% faster than AI Mode queries overall in the last six months. (searchenginejournal.com) Google has been steering toward that behavior for more than a year. In its 2025 AI Mode update, the company said people were asking “more complex, longer and multimodal questions” in Search, and at I/O 2026 it said AI Mode is now powered by Gemini 3.5 Flash by default globally. Google also described the new search box as its biggest upgrade in more than 25 years. (searchenginejournal.com) ### Why does “planning” matter more than a normal search query? Planning queries are harder to answer with a single blue link. Google’s own framing is that users are increasingly asking Search to help compare, narrow, and sequence choices, not just retrieve a fact. That pushes Search toward multi-step exchanges where context from one turn affects the next. (blog.google) Sundar Pichai said at I/O 2026 that Google is in “the agentic Gemini era,” and Google’s Search team said AI Mode is being upgraded for questions that do not “fit into keywords.” Ars Technica reported that Google is moving toward more agentic AI in Search during 2026, meaning the product is being designed for longer tasks rather than one-shot answers. (searchenginejournal.com) ### Where do ads fit into an AI conversation? Google said this week that Gemini-powered ad experiences are being tested inside AI Mode. The company’s ads blog said the formats are designed to offer product guidance inside conversational flows, and it named new formats including Conversational Discovery ads, Highlighted Answers, and AI-powered Shopping ads. Engadget reported that Google is also testing two new ad types tied to the broader AI search experience. (blog.google) That matters because AI Mode keeps users inside a longer interaction. In a conventional results page, ads sit beside or above ranked links. In a conversational interface, ad placement has to fit into an answer stream without obscuring what is sponsored and what is generated. Google said the new formats will include sponsored labeling. (blog.google) ### What does this force Google’s product teams to build? AI Mode requires Search to behave less like a page load and more like a session. Google’s own product updates point to streaming responses, multimodal input, and follow-up turns that preserve context across a conversation. Those features imply backend work for state handling and faster model serving, and frontend work for rendering partial answers, citations, and ad units inside a live exchange. (blog.google) This is an inference from the product capabilities Google described, not a direct company statement. Google’s developer and keynote material also tied Gemini 3.5 Flash to “real-world agentic workflows” and faster response performance. That suggests the company is optimizing Search not just for answer quality but for latency under repeated, multi-turn use. ### What happens next as AI Mode expands? (blog.google) Google said at I/O 2026 that Gemini 3.5 Flash is now the default model in AI Mode globally, and the company’s Search announcements this week paired that rollout with new interface and ad experiments. The next milestones to watch are broader deployment of the ad formats Google previewed and any further product updates on agentic search tasks during 2026. (blog.google 1) (blog.google 2)