New Google Patent May Send Search Traffic to AI Pages Instead of Your Site

The landscape of Search Engine Optimisation (SEO) is shifting under our feet. For decades, the “social contract” between Google and website owners was simple: publishers provide high-quality content, and Google provides traffic via a list of blue links.

However, a recently surfaced Google patent suggests that this contract is being rewritten. As first reported by Google has explored a system that could generate its own AI-powered pages in response to a query, potentially keeping users on Google’s platform rather than sending them to your website.

Here is a deep dive into what this means for the future of digital marketing.

What is the Patent?

The patent, titled “Generating a search results page,” describes a method where Google’s system doesn’t just find the best existing webpage to answer a query—it creates one on the fly.

Instead of a traditional Search Engine Results Page (SERP) that acts as a directory to the web, this system would use Large Language Models (LLMs) to synthesise information from various sources into a cohesive, custom-built landing page. As outlined in the official patent documentation, the goal is to provide a more “resource-efficient” way to answer complex user queries.

Key Features of the Patent:

• Dynamic Page Generation: The AI creates a unique page layout and content specifically tailored to the user’s intent.

• Information Synthesis: It pulls snippets, facts, and data from multiple high-ranking sites to build a comprehensive answer.

• Interactive Elements: The generated page might include AI-driven follow-up questions, tools, or multimedia elements.

Why This Is Different from “AI Overviews”

You might be thinking, “Isn’t this just Search Generative Experience (SGE) or AI Overviews?” Not exactly.

Current AI Overviews sit at the top of the search results, often summarizing content and providing links to sources. The architecture described in suggests a much more radical shift to an AI-generated content page tailored to a specific user.

Rather than a summary followed by links, the AI-generated page becomes the final destination. If the AI page perfectly answers a “How to fix a leaky faucet” query with diagrams and steps synthesized from five different DIY blogs, the user has zero reason to click through to any of those sources.

The “Content Paradox”: Using Your Data to Replace You

The most controversial aspect of this patent for publishers is the source of the information. To generate a high-quality “bespoke” page, Google needs data. That data comes from the “Open Web”—the very blogs, news sites, and niche publishers that rely on Google for traffic.

The cycle looks like this:

1. A publisher writes a high-quality article.

2. Google crawls and indexes that article.

3. The AI “reads” the article and learns the information.

4. A user searches for that topic.

5. Google generates an AI page using the publisher’s information but doesn’t send the user to the publisher’s site.

This creates a “Zero-Click Search” environment on steroids, where the value of the content is extracted by the search engine, leaving the creator with no measurable ROI.

What Types of Content are Most at Risk?

Based on the patent’s focus on “informational synthesis,” certain types of websites are more vulnerable than others:

• Informational/Educational Sites: Sites that provide “what is” or “how-to” content.

• Affiliate Review Sites: If Google can synthesize the “top 10 laptops” based on consensus data, individual review sites may see a sharp decline in clicks.

• Recipe Blogs: If a custom AI page displays the ingredients and steps without the “fluff,” the original site loses the ad impressions.

• Aggregators: Sites that collect data (weather, stock prices, basic lists) are already being replaced.

How Publishers Can Fight Back: The “Un-AI-able” Strategy?

If Google moves toward becoming an “Answer Engine” rather than a “Search Engine,” how do you survive?

1. Focus on E-E-A-T (Experience and Expertise)

AI is excellent at summarizing “common knowledge.” It is terrible at providing first-hand experience. Content that includes personal anecdotes or “Inside my process” is harder for an AI to replace because the value is the person, not just the fact.

2. Build a Direct Brand

If your traffic comes 100% from Google, you are in a precarious position. Diversify through email newsletters, Discord communities, or video platforms where users go for personality and connection.

3. Proprietary Data and Original Research

An AI can summarize existing data, but it cannot conduct an original survey or interview an industry expert. Conducting original research makes you the “Primary Source” that Google (and its AI) must credit.

Conclusion

A patent is not a product launch. Google files thousands of patents every year, and many never see the light of day. However, this patent provides a clear window into Google’s North Star: reducing friction for the user.

For Google, the “perfect” search engine gives the user the answer instantly. For publishers, the “perfect” search engine sends them a visitor. As these two goals drift further apart, the websites that survive will be those that offer a unique human perspective that an algorithm simply cannot replicate.

Is your site ready for the era of AI-generated search pages? Now is the time to start building a brand that users ask for by name.