Google I/O 2026 Rebuilds Search Around Gemini 3.5 Flash, Information Agents, Generative UI

At I/O 2026 Google announced a Search overhaul built on Gemini 3.5 Flash, introduced 'Information Agents' that act on the user's behalf across services, and demonstrated generative UI in which widgets are constructed on-the-fly per query rather than served from fixed templates. The framing from Google is that Search is no longer ten blue links plus an AI summary — it is a generative interface.

Ken Huang's widely-shared analysis (linked from HN) summarised it as Google 'stitching together models, tools, sandboxes, browsers, Search, Workspace, Android, Chrome, and Cloud into one agentic operating layer.' That posture leans on assets no other vendor has — Chrome, Android, Workspace, YouTube — and is Google's most explicit attempt yet to make distribution the moat.

The SEO and publisher communities reacted sharply: generative UI replaces traditional results pages, and CNN's lawsuit against Perplexity the same week reads as a preview of the legal pressure Google will face if generative UI consistently substitutes for clicks. Demis Hassabis amplified demos of Google Omni generating drone-POV footage from sketched camera paths, reinforcing the multimodal frame.

What to watch: rollout cadence of Information Agents to non-US markets, ad-format implications of generative UI, and whether Apple's reported dissatisfaction with ChatGPT integration pushes Siri toward Gemini.