Overview
- Google unveiled the redesign at its I/O conference on May 19, converting the traditional search box into a larger, multimodal entry that accepts long text, images, files, videos and Chrome tabs and supports follow-up conversational sessions.
- The new experience is powered by Gemini 3.5 Flash, which Google says is faster and cheaper to run and will be the default engine behind AI Overviews, AI Mode and new generative UI features.
- Google introduced background, agentic tools including Search agents and a personal agent called Gemini Spark that Google is testing with limited users and plans to make available to paid subscribers in a higher tier.
- The company is rolling out new monetization formats—conversational sponsored explainers, Business Agent for Leads and ads embedded in AI responses—with additional agent actions and Antigravity mini-app integrations slated for broader release this summer.
- Publishers and experts warn the shift deepens the ‘zero-click’ problem and accuracy risks, noting studies that found AI assistants often fail to provide reliable citations and that reduced referral traffic could hurt independent news and niche sites.