The 20-second version
- On June 5, 2026, Google updated its official guide to optimizing for generative AI features on Search.
- Key line: "optimizing for generative AI search is optimizing for the search experience, and thus still SEO."
- A "mythbusting" section declares these unnecessary: llms.txt, content chunking, special schema.org, AI-specific writing, and inauthentic mentions.
- What actually matters: unique, indexable content, a clean technical base, and snippet eligibility.
On June 5, 2026, Google updated its official documentation, "Optimizing for generative AI features on Google Search," published on Search Central. The message fits in one sentence: "From Google Search's perspective, optimizing for generative AI search is optimizing for the search experience, and thus still SEO." In plain terms: for Google, AEO and GEO are not new disciplines, they are SEO.
This is no small claim. For two years, an entire industry of "GEO hacks" has been built on the idea that special techniques are needed to get cited by AI. Google just answered, in black and white, in its own documentation.
What Google declares unnecessary
The most-discussed addition is a "mythbusting" section listing tactics that have no effect on AI Overviews and AI Mode:
- llms.txt and AI files: "You don't need to create new machine-readable files, AI text files, markup, or Markdown" to appear in generative AI search.
- Chunking: there is no requirement to break content into tiny pieces for AI to understand it better.
- AI-specific writing: the systems understand synonyms and general meaning, so there is no need to capture every keyword variation.
- Special schema.org: "Structured data isn't required for generative AI search, and there's no special schema.org markup you need to add."
- Inauthentic mentions: chasing artificial "mentions" across the web isn't as helpful as it might seem.
In other words, several services sold under the "GEO" label target levers that Google itself says do nothing for its AI features.
How Google's AI actually picks its sources
The guide explains the real mechanism, and it is refreshingly clear. AI features run on RAG (retrieval-augmented generation, also called grounding): Google's usual ranking systems retrieve relevant, up-to-date pages from the Search index, then the model relies on those pages to produce an answer with clickable links. A "query fan-out" mechanism runs related queries in parallel to widen the source pool.
The consequence is simple: if your page is not indexed and snippet-eligible, it cannot be cited by AI. AI citation is not a parallel channel, it is a layer on top of the classic index. This echoes what competing engines already do with passage-level retrieval: page quality stays the raw material.
Is your content actually "retrievable" by Google's AI? We check your indexation, your snippet eligibility and your editorial coverage, then tell you where to act first.
What it changes for small businesses
The relief is real: no need to invest in exotic "GEO" infrastructure. But the flip side is just as real. Google sends everyone back to the fundamentals, and the fundamentals are demanding. Here is the useful reading.
| Stop doing | Double down on |
|---|---|
| Buying a "GEO pack" centered on llms.txt / special schema | Non-commodity content with a real point of view |
| Rewriting "to please the AI" | Clean indexation and snippet eligibility |
| Buying artificial mentions | Technical base: semantic HTML, crawl, JS SEO, page experience |
| Spinning up content variations | Up-to-date Merchant Center feeds and Google Business Profile |
For e-commerce and local, Google highlights two concrete signals: well-maintained Merchant Center feeds and an up-to-date Google Business Profile. Nothing flashy, but this is the data that feeds AI answers for commercial and local queries. The broader context stays tense: AI Overviews now trigger on a majority of business queries, which makes citation eligibility decisive.
Limits: what this guide does not say
This documentation is about Google only. ChatGPT Search, Perplexity and Claude have neither the same index nor the same citation rules. On those engines, presence on third-party sources (Reddit, Wikipedia, media) and passage structure matter more. "GEO is dead" therefore only holds for the Google ecosystem.
The guide also gives no citation guarantee: being indexable is necessary, not sufficient. And it does not quantify the impact of AI Overviews on traffic, a topic we covered alongside the end of the May 2026 Core Update and the arrival of AI reports in Search Console.
The Cicéro take
This guide validates what we have repeated since AI Overviews launched: the best GEO is serious SEO applied to content the AI cannot fabricate on its own. No magic file, no miracle tag. Useful, indexed, technically clean content with an angle. If a vendor sells you a "GEO pack" built on llms.txt, show them Google's page.
Sources
- → Google Search Central — "Optimizing for generative AI features on Google Search" (official documentation, updated 2026-06-05)
- → Search Engine Journal — "Google's New AI Search Guide Calls AEO And GEO 'Still SEO'"
Frequently asked questions
Do you need an llms.txt file to appear in Google's AI answers?
Are GEO and AEO different from SEO according to Google?
Do AI Overviews and AI Mode require special schema.org markup?
How does Google's AI pick the pages it cites?
Growth and SEO content strategist, I founded Cicéro to help businesses build lasting organic visibility — on Google and in AI-generated answers alike. Every piece of content we produce is designed to convert, not just to exist.
LinkedIn