On May 15, 2026, Google published something the search industry had been waiting for since AI Overviews launched: an official, first-party guide to optimizing for generative AI features in Google Search. The document, titled Optimizing your website for generative AI features on Google Search, is the most important piece of Google documentation for the GEO category to date. Not because it reveals new techniques, but because it systematically dismantles several of them.
On the same day, Google updated its spam policies to explicitly state that all spam rules apply to "generative AI responses in Google Search," including AI Overviews and AI Mode. The message is deliberate: Google is drawing a line between legitimate optimization and manipulation, and it is doing it three days before Google I/O.
This article breaks down what Google confirmed, what it debunked, and why the full picture for brands is more complicated than a single Google guide suggests.
What Google Confirmed: RAG, Fan-Out, and SEO Fundamentals
The guide opens with a direct answer to the question every SEO professional has been asking: is SEO still relevant for AI search? Google's answer is unequivocal. "In short, yes!"
The technical explanation matters here. Google confirms that its generative AI features rely on two core techniques:
Retrieval-augmented generation (RAG): Google's AI uses its core Search ranking systems to retrieve relevant, up-to-date web pages from the Search index. It then reviews information from those retrieved pages to generate a response, showing "prominent, clickable links to relevant web pages that support the information in the response." Query fan-out: The model generates a set of concurrent, related queries to fetch additional relevant search results. For example, a query about fixing a lawn full of weeds might trigger fan-out queries like "best herbicides for lawns," "remove weeds without chemicals," and "how to prevent weeds in lawn."This is structurally significant. RAG means your content must already rank in Google's core Search index to appear in AI responses. There is no separate "AI index" or "AI crawl." Google's AI pulls from the same ranked results that power traditional search. Fan-out means Google is querying multiple angles of a topic simultaneously, so comprehensive content that addresses related subtopics has an advantage over narrowly focused pages.
The implication is clear: if your SEO is weak, your AI visibility on Google will also be weak. The two systems share the same foundation.
Google also confirms that all existing technical SEO best practices remain in force. Pages must be indexed and eligible for snippets. JavaScript must follow established SEO guidelines. Semantic HTML helps but does not need to be perfect. Crawlability remains essential. Page experience matters. Duplicate content should be reduced. None of this is new, but having it stated in the context of AI optimization gives it fresh authority.
The Number One Recommendation: Non-Commodity Content
Google's highest-leverage recommendation is not technical. It is editorial. The guide states that "creating content that people find unique, compelling, and useful will likely influence your website's presence in generative AI search in the long run more than any of the other suggestions in this guide."
Google draws a sharp distinction between commodity and non-commodity content:
- Commodity content: "7 Tips for First-Time Homebuyers." Based on common knowledge. Could originate from anyone. Adds little unique insight.
- Non-commodity content: "Why We Waived the Inspection and Saved Money: A Look Inside the Sewer Line." Unique expert or experienced takes that go beyond common knowledge and the ordinary.
This distinction is the heart of the guide's content philosophy. Google is telling publishers that its AI systems look at "a variety of sources," and content with a "unique viewpoint that stands out" has a structural advantage. First-hand experience beats recycled summaries. Original analysis beats AI-generated rewrites of existing content. Expert depth beats broad-but-shallow coverage.
For brands, this is both guidance and warning. If your content strategy relies on producing high volumes of commodity content, your AI visibility on Google is going to suffer regardless of any technical optimization you apply.
The Debunking: Five Things Google Says You Do Not Need
The most provocative section of the guide is titled "Mythbusting generative AI search: what you don't need to do." Google addresses five specific tactics that have become popular in the GEO industry, and dismisses each one.
1. You do not need llms.txt files
Google's language is direct: "You don't need to create new machine readable files, AI text files, markup, or Markdown to appear in generative AI search." The guide adds that while Google may discover and crawl many file types on a website, "this doesn't mean that the file is treated in a special way."
This is a direct hit at one of the most widely promoted GEO tactics. llms.txt, a standardized file that tells AI agents how to access a site's content, has been adopted by brands and recommended by consultants as a foundational AI visibility action. Google is saying it does not care about llms.txt for Google Search AI features. Period.
2. You do not need to "chunk" content
"There's no requirement to break your content into tiny pieces for AI to better understand it." Google says its systems can "understand the nuance of multiple topics on a page and show the relevant piece to users." This debunks the common advice to restructure content into small, discrete answer units specifically for AI extraction.
3. You do not need to rewrite content for AI systems
"You don't need to write in a specific way just for generative AI search." Google's AI can understand synonyms and general meanings, connecting users with content that might not use the same precise words as the query. You do not need to worry about capturing every long-tail variation.
4. You do not need inauthentic "mentions"
"Just like the rest of Google Search, our generative AI features can show what's being said about products and services across the web." But seeking inauthentic mentions "isn't as helpful as it might seem." Google's core ranking systems focus on high-quality content, and its spam systems block manipulation. AI features depend on both.
5. You should not overfocus on structured data
"Structured data isn't required for generative AI search, and there's no special schema.org markup you need to add." Google recommends continuing to use structured data as part of an overall SEO strategy for rich results, but it is not a special lever for AI visibility.
Google also warns explicitly against creating separate pages for every fan-out query variation. Doing so "primarily to manipulate rankings or generative AI responses in Google Search violates Google's scaled content abuse spam policy."
The Enforcement: Spam Policies Now Cover AI Responses
The guide did not arrive alone. On the same day, Google updated the lead paragraph of its spam policies to clarify that all spam rules apply to "generative AI responses in Google Search." The updated language reads: "spam refers to techniques used to deceive users or manipulate our Search systems into featuring content prominently, such as attempting to manipulate Search systems into ranking content highly or attempting to manipulate generative AI responses in Google Search."
Search Engine Roundtable reported that the change was made to make "it clear that the spam policies apply to all of Google Search, including generative AI responses."This is not theoretical. Any tactic designed specifically to manipulate AI Overviews or AI Mode responses, from fake mentions to scaled content targeting fan-out queries, is now formally classified as spam. Google has given itself the enforcement mechanism to penalize sites that treat AI features as an optimization surface separate from core Search quality.
Agentic Experiences: Google Acknowledges Browser Agents
The guide includes a notable section on "agentic experiences." Google defines AI agents as "autonomous systems that can perform tasks on behalf of people, such as booking a reservation or comparing product specifications." It acknowledges that "browser agents may access your website to gather the data they need to complete these tasks, such as analyzing visual renderings (like screenshots), inspecting the DOM structure, and interpreting the accessibility tree."
This is Google's first official acknowledgment that browser-based AI agents are a meaningful category for website owners to consider. The guidance is minimal for now: focus on semantic HTML, ensure content is accessible through the DOM, and maintain a rendering-friendly site. But the inclusion signals that Google sees agentic commerce as a real, near-term consideration, not a theoretical future.
Why This Guide Matters More Than It Seems
This is not just a documentation update. It is a strategic positioning move by Google, and it has implications that go well beyond SEO.
First, it sets the frame for Google I/O, which starts May 19. By publishing this guide three days before the event, Google is establishing its official position on AI optimization before any announcements that might expand AI Mode, deepen Gemini integration, or change how AI features work in Search. Whatever Google announces at I/O, this guide is the baseline.
Second, it is a direct response to the growing GEO industry. Terms like "GEO" and "AEO" (Answer Engine Optimization) have become common in marketing circles. Google is pushing back on the framing that AI optimization requires a fundamentally different approach than SEO. The guide explicitly states that "many suggested 'hacks' aren't effective or supported by how Google Search actually works."
Third, it gives Google enforcement cover. By codifying what does not work and updating spam policies in the same update, Google is creating a clearer boundary between legitimate optimization and manipulation. Sites that pursue debunked tactics now have less plausible deniability.
The Nuanced Reality: Google Is One Engine Among Many
Here is where the guide's scope limitation becomes critical. Google's document is about Google Search AI features only: AI Overviews and AI Mode. It is authoritative for Google. But Google is one of at least four major AI discovery surfaces that brands need to care about, and the rules for the other three are meaningfully different.
ChatGPT uses Bing's index combined with its own training data. It does not rely on RAG from Google's ranking systems. Content that ranks poorly in Google can still be cited by ChatGPT if it ranks well in Bing or is present in ChatGPT's training corpus. ChatGPT also actively consumes llms.txt files from websites, something Google explicitly says it ignores. Perplexity operates its own independent crawling and indexing system. It uses its own ranking signals, not Google's. Perplexity has been one of the strongest proponents of llms.txt adoption, and structured, answer-first content has a measurable citation advantage on Perplexity. Google's "you don't need to write in a specific way" advice does not apply to Perplexity's source-selection mechanics. Claude uses a combination of web search results and its own knowledge base. It has different citation patterns than both Google and ChatGPT, with a stronger preference for well-structured, source-heavy content with clear definitions.The practical implication is straightforward. If you optimize only for Google's AI features using only Google's guidelines, you will perform well on Google but may underperform on ChatGPT, Perplexity, and Claude. Google's guide is the floor, not the ceiling. A comprehensive GEO strategy accounts for the source-selection mechanics of all major AI engines, not just Google.
What Smart Brands Should Actually Do
Google's guide is right about one thing: strong SEO fundamentals are the foundation. If your content does not rank, it will not appear in Google's AI responses. That is not negotiable. But there is more to the picture.
Follow Google's guide for Google. Create non-commodity content with a unique point of view. Maintain technical SEO excellence. Do not create spam or scaled content targeting fan-out queries. This will serve you well for AI Overviews and AI Mode. Optimize beyond Google. llms.txt may not matter for Google, but it matters for ChatGPT and Perplexity. Structured data may not be required for Google's AI features, but it helps with rich results across all search engines and improves machine readability for AI agents. Answer-first content may not be necessary for Google, but it demonstrably improves citation rates on Perplexity and Claude. Think multi-engine from the start. The GEO vs SEO comparison is real, but the answer is not to choose one or the other. Good SEO is the foundation for Google's AI features. Good GEO extends that foundation to the other AI engines that an increasing share of your audience is using. Prepare for agents. Google's own guide acknowledges browser agents. The agentic commerce wave is accelerating. Ensure your content is accessible through the DOM, your semantic HTML is clean, and your product and business data is machine-readable. Use the right measurement. Google Analytics now includes an AI Assistant channel group that automatically separates AI chatbot traffic from other channels. Use it. Track your visibility across engines, not just Google. The brands that measure AI visibility comprehensively will be the ones that catch shifts early.The Strategic Takeaway
Google's official AI optimization guide is an important document. It confirms that Google's AI features are built on top of existing Search infrastructure, not alongside it. It debunks several popular GEO tactics with authority that only a first-party source can provide. And it draws a clear enforcement line by updating spam policies in the same move.
But the guide describes Google's engine. The AI discovery landscape in 2026 spans at least four major platforms with meaningfully different source-selection mechanics. Brands that treat Google's guide as the complete playbook for AI visibility will optimize for one engine and miss the other three.
The right approach is simpler than it sounds: build on strong SEO fundamentals, then extend with GEO tactics that work across engines. Non-commodity content wins everywhere. Technical SEO excellence wins everywhere. But llms.txt, structured data, answer-first content, and multi-engine monitoring are the extensions that separate a Google-only strategy from a comprehensive AI visibility strategy.
Google's guide is the floor. Build from there.
---
Is your brand visible across all four major AI engines, or just Google? Run a comprehensive AI visibility audit and find out where you stand. Check your AI visibility now.---
Sources
1. Google Developers. "Optimizing your website for generative AI features on Google Search." May 15, 2026. developers.google.com/search/docs/fundamentals/ai-optimization-guide
2. Google Developers. "Google Search spam policies." Updated May 15, 2026. developers.google.com/search/docs/essentials/spam-policies
3. Schwartz, Barry. "Google: Spam Policies Apply To AI Responses (AI Overviews & AI Mode)." Search Engine Roundtable. May 15, 2026. seroundtable.com/google-spam-policies-apply-ai-responses-41331.html
4. Schwartz, Barry. "Google publishes guide on optimizing for generative AI features." Search Engine Land. May 15, 2026. searchengineland.com/google-publishes-guide-on-optimizing-for-generative-ai-features-477671
5. Gizmodo. "Google's Spam Policies Now Apply to Attempts to Manipulate AI." May 15, 2026. gizmodo.com/googles-spam-policies-now-apply-to-attempts-to-manipulate-ai-2000759393
6. The Searchless Journal. "How ChatGPT Chooses Sources: AI Citation Mechanics for 2026." May 11, 2026. searchless.ai/articles/2026-05-11-how-chatgpt-chooses-sources-citation-mechanics-2026/
7. The Searchless Journal. "How Perplexity Chooses Sources: Citation Analysis and Optimization." May 9, 2026. searchless.ai/articles/2026-05-09-how-perplexity-chooses-sources-citation-mechanics-2026/
8. The Searchless Journal. "GEO vs SEO: Complete Comparison and Migration Strategy for 2026." May 15, 2026. searchless.ai/articles/2026-05-15-geo-vs-seo-complete-comparison-migration-strategy-2026/
---
FAQ
Does Google's guide mean GEO is dead?
No. Google's guide says that for Google Search AI features (AI Overviews, AI Mode), strong SEO fundamentals are the primary optimization path, and several popular "GEO hacks" do not work on Google. But GEO as a category covers optimization across all AI engines, not just Google. ChatGPT, Perplexity, and Claude have different source-selection mechanics where tactics like llms.txt, answer-first content, and structured data do make a measurable difference.
Should I still implement llms.txt?
Yes. Google says llms.txt does not matter for Google Search AI features. But ChatGPT and Perplexity actively consume llms.txt files, and these platforms represent a growing share of AI-driven discovery. Implementing llms.txt is a low-effort action that improves visibility on the engines where it matters, without harming your Google presence.
What does "non-commodity content" mean in practice?
Google defines commodity content as information based on common knowledge that could come from anyone. Think generic listicles and basic definitions. Non-commodity content provides unique perspective, first-hand experience, or expert analysis that goes beyond what is already available. In practice, this means writing from direct experience, sharing original data or analysis, and offering opinions backed by evidence rather than restating conventional wisdom.
Does this guide apply to ChatGPT, Perplexity, and Claude?
No. The guide is specifically about Google Search AI features. Each AI engine has its own source-selection mechanics, ranking signals, and content preferences. A strategy optimized purely for Google's AI features will not necessarily perform well on other AI platforms.
What changed with Google's spam policies?
Google updated the opening paragraph of its spam policies to explicitly state that spam rules apply to "generative AI responses in Google Search," including AI Overviews and AI Mode. This means any tactic designed specifically to manipulate AI-generated responses, from fake mentions to scaled content targeting fan-out queries, is now formally classified as spam and can trigger enforcement action.
---
Learn more about AI visibility measurement and methodology at searchless.ai/ai-visibility.