AEO and GEO Are Still SEO: What Google's Official AI Guide Actually Says

Google finally said it out loud

For two years, an entire cottage industry grew up around a single assumption: that AI search needed its own playbook. Answer Engine Optimization. Generative Engine Optimization. A wave of "hacks": special text files, content chopped into bite-size chunks, copy rewritten to sound machine-friendly.

On May 15, 2026, Google published its first official guidance on the subject and answered the question directly. "Is SEO still relevant for generative AI search?" the document asks. Its answer: "In short, yes!"

That single line reframes the last two years of debate. Google's generative AI features, AI Overviews and AI Mode, are in its own words "rooted in our core Search ranking and quality systems." The same index. The same quality signals. The same fundamentals.

AEO is still SEO. GEO is still SEO. The work that earns visibility in an AI answer is the work that earned visibility in a blue link. What changed is not the discipline. It's the surface. The trade press read it the same way: Search Engine Journal's headline put it plainly, calling AEO and GEO "still SEO." If you're just getting started, here's what to do first when you are tasked with GEO.

But before you throw out everything you've read about optimizing for AI, there's a critical boundary in this guide that almost every hot take has missed. It's the difference between advice that compounds and advice that quietly misleads you. We'll get to it.

How Google's AI features actually pull your content

The guide itself is unusually specific about the machinery, and it's worth understanding because it explains why the fundamentals still rule.

Two systems do the work:

Retrieval-augmented generation (RAG). Google calls this "grounding." Rather than inventing an answer, the model leans on Google's "core Search ranking systems to retrieve relevant, up-to-date web pages" from the index, then synthesizes a response with "prominent, clickable links to relevant web pages that support the information." If your page can't be retrieved by ranking systems, it can't ground an answer.

Query fan-out. When someone asks a question, the model spins up "a set of concurrent, related queries" to gather more context. Google's own example: a search for "how to fix a lawn that's full of weeds" fans out into "best herbicides for lawns," "remove weeds without chemicals," and "how to prevent weeds in lawn."

Read those two mechanisms together and the strategy writes itself. Visibility in AI answers is downstream of being retrievable, rankable, and genuinely relevant across a cluster of related questions, not a single keyword. That's not a new game. It's SEO with a wider aperture.

The five "AI hacks" Google just told you to ignore

The most quoted part of the guide is its mythbusting section, titled, bluntly, "what you don't need to do." Google names the popular tactics directly and tells you to stop. There are five of them, and Search Engine Land's rundown of the guide flags the same list.

1. llms.txt and other "special" markup. "You don't need to create new machine readable files, AI text files, markup, or Markdown to appear in generative AI search," the guide states. Google may crawl such a file like any other, but "this doesn't mean that the file is treated in a special way." One of 2025's most-marketed GEO tactics, declared inert.

2. "Chunking" content. "There's no requirement to break your content into tiny pieces for AI to better understand it." Google's systems "understand the nuance of multiple topics on a page." There is, in Google's words, "no ideal page length." Write for the audience, not the parser.

3. Rewriting content just for AI. "You don't need to write in a specific way just for generative AI search." The systems understand synonyms and intent, so you don't have to stuff in every long-tail variation or contort your prose into "AI-friendly" phrasing.

4. Chasing inauthentic "mentions." Manufacturing brand mentions across the web "isn't as helpful as it might seem." Google's ranking systems reward high-quality content and its spam systems filter the rest, and its AI features depend on both.

5. Overfocusing on structured data. This is the one the takes get wrong, so read it carefully. Structured data "isn't required for generative AI search, and there's no special schema.org markup you need to add." But this part matters: "it's a good idea to continue using it as part of your overall SEO strategy, as it helps with being eligible for rich results on Google Search."

Google did not kill structured data. It killed the idea that you bolt on schema to win AI citations. Keep your FAQ and how-to markup for the rich results it has always earned. Just stop treating it as an AI cheat code. (If you've read our guide to FAQ schema for AI search, this is the nuance to hold onto: the schema still works, for the job it was always doing.)

The three things Google says actually move the needle

Strip out the hacks and the guide is left with a short, almost old-fashioned list. Three priorities:

Create valuable, non-commodity content. Google says this "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." The distinction it draws is sharp. Commodity content (its example: "7 Tips for First-Time Homebuyers") restates common knowledge anyone could produce. Non-commodity content (its example: "Why We Waived the Inspection & Saved Money: A Look Inside the Sewer Line") carries a first-hand, expert point of view "that go[es] beyond common knowledge." The instruction is explicit: "Don't just recycle what others on the internet have already said, or could easily be produced by a generative AI model."

Build and maintain clear technical structure. To appear in AI features, a page "must be indexed and eligible to be shown in Google Search with a snippet." Crawlable content, sound JavaScript handling, good page experience, low duplication. Nothing exotic: the technical SEO checklist you already know.

Optimize your business details. For local and ecommerce, keeping Merchant Center feeds and Google Business Profile current helps products and services surface in AI responses. Structured operational data, kept fresh.

Unique perspective. Clean structure. Accurate, current details. That's the whole list.

The boundary nobody is talking about

Here's the line that separates durable advice from the half-true version going viral: this guide is about Google.

Every word of it governs Google's own generative surfaces, AI Overviews and AI Mode, drawing from Google's index. It says nothing about how ChatGPT, Perplexity, Claude, or Gemini's standalone app decide what to cite. Those engines run their own retrieval, their own freshness windows, their own source preferences, and they don't always land on the same pages Google does.

So when Google says llms.txt does nothing in Google Search, that is true and authoritative, for Google Search. It is not a universal verdict on every AI engine that crawls the web. Conflating the two is how a correct Google statement becomes a misleading industry headline.

This is the practical takeaway: the fundamentals Google endorses (genuinely useful content, clean structure, technical accessibility) are the things that travel well across every engine, because every engine is trying to surface the same thing. The hacks it dismisses were always fragile precisely because they optimized for a parser instead of a person. Do the durable work and you compound everywhere. Chase the hack and you were never going to last anywhere.

But "the fundamentals compound across engines" is a strategy, not a dashboard. You still can't manage what you can't see, and no single search console tells you whether ChatGPT, Perplexity, and Google's AI are each citing you, or quietly citing a competitor instead.

What to actually do now

If you spent the last year buying AI-specific tactics, the guide is less an indictment than a relief. The expensive, fiddly stuff was optional. The work that matters is work you already understand.

A durable workflow looks like this:

• Research the real question cluster. Query fan-out means you're competing across a web of related intents, not one keyword. Map the cluster, not the phrase.
• Write the non-commodity version. First-hand experience, a genuine point of view, the thing a generative model can't produce by recombining what already exists. This is the single highest-leverage input Google names.
• Get the structure right. Indexable, crawlable, fast, well-organized. Keep your schema for rich results; don't worship it.
• Monitor across engines, not just Google. Track whether you're actually being cited in AI Overviews and in ChatGPT, Perplexity, Claude, and the rest. Google's console won't show you the others. If you want to see why a page can rank on Google yet still go uncited by AI, we broke that gap down in why your page ranks but isn't cited by AI.
• Fix what's slipping, and republish. Citations decay as engines refresh and competitors update. Visibility is a maintenance loop, not a launch.

That loop (research, write, optimize, monitor, fix, republish) is exactly the "still SEO" discipline Google just validated, extended across every AI surface instead of one. It's also what Frase is built to run end to end: the research and writing, the optimization, and the part most tools skip, watching your visibility across AI engines and closing the gap when it opens.

You don't need a secret AI file. You need to do the real work, see where it's landing across the engines that matter, and keep it current. Google just confirmed the first part. The second part is why visibility monitoring exists.

Want to see who's actually citing you across AI engines, not just Google? Check your AI visibility free and find out where you show up, and where you don't.

Frequently asked questions

Is SEO still relevant for AI search?

Yes. Google's May 2026 guide states its generative AI features, AI Overviews and AI Mode, are "rooted in our core Search ranking and quality systems." The same index and quality signals that earn rankings are what earn citations in AI answers, so SEO is still the discipline that drives AI visibility on Google.

Does Google use llms.txt?

No. Google's guide states you don't need to create machine-readable files, AI text files, or special markup to appear in generative AI search. Google may crawl an llms.txt file like any other page, but "this doesn't mean that the file is treated in a special way" in Google Search.

Do I need to break my content into chunks for AI?

No. Google says there's "no requirement to break your content into tiny pieces" and "no ideal page length." Its systems understand multiple topics on a single page, so you should structure content for readers rather than chopping it up for a parser.

Did Google say to stop using structured data?

No. Google says schema "isn't required for generative AI search" and there's no special markup to add, but it's "a good idea to continue using it" because it helps you stay eligible for rich results in Google Search. Keep your schema; just don't treat it as an AI citation hack.

What actually improves AI search visibility on Google?

Three things: valuable, non-commodity content with first-hand expertise; clean, indexable technical structure; and accurate, current business details. Google calls non-commodity content the factor most likely to influence your presence in generative AI search over the long run.

Does Google's AI guide apply to ChatGPT and Perplexity?

No. The guide governs Google's own AI surfaces, AI Overviews and AI Mode, which draw from Google's index. ChatGPT, Perplexity, Claude, and Gemini's app run their own retrieval and source preferences, so "llms.txt does nothing" is true for Google Search but not a universal verdict for every AI engine.

What is query fan-out in Google's AI search?

Query fan-out is when Google's model generates "a set of concurrent, related queries" to gather context for an answer. It means your visibility depends on being relevant across a cluster of related questions, not just one keyword, so plan content around the full question cluster.

How do I monitor whether AI engines are citing me?

Google Search Console only reports Google. To see whether ChatGPT, Perplexity, Claude, and Google's AI features are citing you (or a competitor instead), use a cross-engine AI visibility tool. Frase's free AI Visibility Checker shows where you show up across engines and where you don't.