Google Is Rebuilding Search From Scratch — And It Looks Nothing Like Search
In a nutshell
For 28 years, Google Search has worked the same way. You type a query. Google returns a list of links. You click one. You leave Google. The publisher gets a visit. The advertiser gets an impression.
That model is ending.
What Google Is Building
Google is rebuilding the search bar itself. A new feature lets Google generate custom visuals, interactive graphics, and even mini-apps directly on the search page in response to queries — piecing together an answer that requires no click, no link and no departure from Google's own environment.
This is not AI Overviews — the text summaries that already appear above search results and have already reduced click-through rates by up to 61%. This is something more fundamental: Google generating functional interactive content — graphics, calculators, comparisons, mini-applications — that answers queries completely within the search interface.
A user searching for "best laptop under €1000" would not see a list of tech review links. They would see an interactive comparison tool — generated by Google, powered by Gemini, displaying specifications, prices and recommendations — without a single external website involved.
The Publisher Extinction Event
We have reported consistently on the traffic impact of AI Overviews on European publishers. What Google is now building makes AI Overviews look like a minor inconvenience.
If Google can generate interactive graphics and mini-apps in response to queries — content that previously required visiting specialised websites — the business model of every informational website on the internet faces an existential challenge. Recipe sites. Comparison tools. Financial calculators. Travel guides. Educational content. Product reviews.
All of these categories can be replaced by Google-generated interactive content that never sends traffic to the original creator.
Adobe, Canva, and CapCut announced Gemini integrations to let users access their image and video editing tools directly inside the Gemini app. Users can generate AI content in Gemini and refine it in professional editors without switching applications.
The pattern is consistent across every Google product announced in the past two weeks: bring everything inside Google's environment. Search. Gemini. Maps. YouTube. Commerce. The open web — where users leave Google to visit other destinations — is being replaced by a closed Google ecosystem where everything happens within Google's walls.
The Chinese AI Dimension — The Story Nobody Is Watching
The most significant development in the global AI landscape this week has nothing to do with GAFAM. But it has everything to do with Google's competitive position.
Chinese models — Kimi K2.6, DeepSeek V4, GLM-5.1, Qwen 3 — now account for 60% of all AI usage on OpenRouter, the most-used third-party AI model router. This is the clearest single signal that the open-weights tier is now Chinese-led. Meta's delayed Avocado model — the last credible US open-weights frontier candidate — has gone silent.
Sixty percent. Chinese models. On the world's most-used AI model router.
This data point — buried in this week's AI coverage — is the most important competitive intelligence signal of 2026 for every GAFAM company. The open-source AI tier that Meta was supposed to lead — and that gave Western developers an alternative to proprietary models — is now dominated by Chinese models that the US government cannot sanction, that European regulators cannot easily scrutinise and that GAFAM companies cannot compete with on price.
Google's response — rebuilding Search as an AI operating system, locking users inside its ecosystem — is partly a response to ChatGPT and partly a response to the Chinese open-source threat. If developers can access Chinese models at a fraction of the cost of Gemini, Google's API business faces pressure it has not yet acknowledged publicly.
What This Means for GAFAM
At Google I/O 2026, Google showed what happens when a company with Search, Android, YouTube, Chrome, Workspace, Cloud, and billions of users decides to connect everything with one AI layer. Earlier versions of Gemini felt like a competitor to ChatGPT. Now it feels more like an AI brain sitting inside Google's entire product ecosystem.
That framing — Gemini as an AI brain inside Google's ecosystem rather than a competitor to ChatGPT — is the most important strategic reframe of the AI era. Google is not trying to win the chatbot war. It is trying to make the chatbot war irrelevant by building something the chatbot model cannot compete with: a vertically integrated AI operating system with two and a half billion Android users, the world's most-used search engine and the world's most-used video platform all connected through a single intelligence layer.
The European Perspective
Google's Search rebuild — generating mini-apps and interactive graphics that eliminate the need to visit external websites — will face its most significant regulatory challenge in Europe. The EU's Digital Markets Act requires Google, as a designated gatekeeper, to provide fair access to third-party content and services. Generating AI content that replaces third-party websites — without compensation, without traffic, without attribution — is precisely the gatekeeping behaviour the DMA was designed to prevent. The European Commission's Digital Markets Act enforcement team will be examining Google's Search rebuild with particular attention. The question Brussels must answer: does generating an interactive calculator that replaces a publisher's calculator constitute unfair self-preferencing — or legitimate product innovation? The answer will define the boundary of AI-generated content in European search for a decade. gafam.ai will be watching.
We are not first. We are right.
SOURCES
— Build Fast With AI: AI News Today May 25 2026 — Top AI Stories & Headlines
— Medium / Let's Code Future: Google Just Changed the AI Race Again
— Fortune: Microsoft, Meta, and Google just announced billions more in AI spending