Skip to content

What is AI SEO? How ChatGPT, Google AI Overviews and Perplexity choose their sources

By the Growli teamUpdated 7 min read

Short answer

AI SEO is the work of making your website easy for AI search tools such as ChatGPT search, Google AI Overviews, Perplexity and Copilot to crawl, understand and cite. It builds on classic SEO: crawlable pages, indexing in Google and Bing, clear answers, consistent facts about your business and solid technical basics.

Where to start

  1. Step 1: Let AI crawlers reach your site

    Check robots.txt and your firewall so search bots such as OAI-SearchBot, Claude-SearchBot and PerplexityBot are not blocked.

  2. Step 2: Get indexed in Google and Bing

    Many AI answers start from search results. Use Google Search Console and Bing Webmaster Tools.

  3. Step 3: Put a direct answer at the top of key pages

    Say who you are, what you offer, for whom and where in the first lines.

  4. Step 4: Add Organization structured data

    It states your name, logo and profiles in a format machines read reliably.

  5. Step 5: Check where you stand

    A free Growli audit checks AI crawler access, structured data, llms.txt and more in about a minute.

    Run a free audit

What does AI SEO actually mean?

AI SEO means optimizing your site so that AI assistants and AI search features can find your pages, understand them and use them as sources in their answers. You will also see it called GEO (generative engine optimization), AEO (answer engine optimization) or AI search optimization. The names differ, the goal is the same.

Classic SEO aims for a good position in a list of ten blue links. AI search often gives the user a written answer first, with a handful of linked sources beside or below it. Being one of those sources, or being named in the answer, is what AI SEO is about.

It is not a separate discipline that replaces SEO. Most AI search tools still rely on a web index and on crawlers, so the same foundations decide whether you are even in the running.

How do AI search tools find and pick sources?

Each tool works a little differently, but the pattern is similar. A crawler or a search index supplies candidate pages, the system retrieves the ones that look relevant to the question, and a language model writes an answer that cites some of them. The details below come from each company's own documentation and can change, so check the linked pages when it matters.

ChatGPT search

OpenAI uses a crawler called OAI-SearchBot to surface websites in ChatGPT's search features, and it recommends allowing it in robots.txt if you want to appear. GPTBot is a separate crawler for model training, and ChatGPT-User fetches pages when a user asks ChatGPT to visit them. OpenAI has said ChatGPT search also uses third-party search providers, without publishing a full breakdown. See OpenAI's crawler overview.

Google AI Overviews, AI Mode and Gemini

Google's AI features in Search use the normal Google index. Google states that a page only needs to be indexed and eligible to show in Search with a snippet, and that there are no extra technical requirements, special schema or AI text files needed. Both features may run several related searches behind the scenes (Google calls this query fan-out) to find supporting pages. See Google's guidance on AI features.

The Google-Extended robots.txt token controls whether your content is used for Gemini model training and grounding. Google says it does not affect inclusion or ranking in Google Search.

Perplexity

Perplexity runs PerplexityBot to surface and link websites in its results and says it is not used to train foundation models. A second agent, Perplexity-User, fetches pages when a user asks a question and generally ignores robots.txt because a person triggered it. See Perplexity's crawler documentation.

Microsoft Copilot and Bing

Copilot is grounded in Bing's index. The Bing Webmaster Guidelines cover Bing search, Copilot and grounding results together, and they state plainly that no optimization guarantees a citation. Bing Webmaster Tools also has an AI Performance report that shows which of your pages are cited.

How is AI SEO different from classic SEO?

The overlap is large: if Google and Bing cannot crawl and index a page, most AI tools will not cite it either. The differences are mostly about what the result looks like and which extra crawlers you need to think about.

  • Topic
    What the user sees
    Classic SEO
    A ranked list of links
    AI SEO
    A written answer with a few cited sources
  • Topic
    Crawlers to allow
    Classic SEO
    Googlebot, Bingbot
    AI SEO
    Also OAI-SearchBot, PerplexityBot, Claude-SearchBot and similar
  • Topic
    Unit of value
    Classic SEO
    A whole page ranking for a query
    AI SEO
    A passage, fact or definition the model can quote
  • Topic
    Success measure
    Classic SEO
    Rankings, clicks, impressions
    AI SEO
    Citations and mentions, plus referral visits
  • Topic
    Reporting
    Classic SEO
    Search Console, Bing Webmaster Tools
    AI SEO
    Partly in the same tools; often incomplete

One practical consequence: a page can rank well and still never be quoted, because the useful answer is buried under a long intro. Another is that AI answers often draw on several sources for one question, so a clear section on a narrower subtopic can earn a citation even when your page is not the top result.

Lever 1: let the right crawlers in

Start with access. Check your robots.txt and your firewall or CDN bot settings, because a security rule that blocks unknown bots can quietly shut out AI crawlers even when robots.txt allows them. Search crawlers and training crawlers are separate, so you can allow one and block the other.

robots.txt
# Allow AI search crawlers
User-agent: OAI-SearchBot
Allow: /

User-agent: PerplexityBot
Allow: /

User-agent: Claude-SearchBot
Allow: /

# Opt out of OpenAI model training (optional)
User-agent: GPTBot
Disallow: /

Sitemap: https://www.example.com/sitemap.xml

Whether to block training crawlers is a business decision, not an SEO requirement. Our guide to robots.txt for AI crawlers lists the main user agents and what each one does.

Lever 2: write answers a model can quote

AI answers are built from passages. Pages that state the answer early and plainly are easier to use than pages that circle the topic. This is also what Google's guidance on helpful content has asked for all along.

  • Put a short, direct answer near the top of the page, then expand on it.
  • Use headings that match real questions, such as "How much does a roof inspection cost?"
  • Keep paragraphs short and give one idea per paragraph.
  • Use lists and tables for steps, options and comparisons.
  • State specific facts (prices, opening hours, service areas, limits) in text, not only in images.
  • Keep pages current and show when they were last updated.

Read your first paragraph on its own. If it does not answer the question the page is about, rewrite it before anything else.

Lever 3: make your business a clear entity

An entity is a thing a search system can identify unambiguously: your company, a product, a person. AI tools describe businesses more accurately when the same name, address, contact details and description appear on your site and on your profiles elsewhere.

Structured data helps with this. It is not required for Google's AI features, but it states your facts in a machine-readable way. A basic Organization block in JSON-LD (a script format for structured data) looks like this:

JSON-LD
<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Example Plumbing Ltd",
  "url": "https://www.example.com",
  "logo": "https://www.example.com/logo.png",
  "telephone": "+44 20 7946 0000",
  "sameAs": [
    "https://www.linkedin.com/company/example-plumbing",
    "https://www.facebook.com/exampleplumbing"
  ]
}
</script>

The schema markup guide covers which types are worth adding for different kinds of sites. Being mentioned and reviewed on other trusted sites also shapes how AI tools see your business; that side is covered in how to get recommended by ChatGPT.

Lever 4: get the technical basics right

Not every AI crawler renders JavaScript the way Googlebot or a browser does. If your text, prices or product details only appear after scripts load, a crawler may see an almost empty page. Server-side rendering or static HTML avoids this, as explained in JavaScript SEO for AI crawlers.

  1. Make sure important pages return a 200 status and are not blocked by noindex.
  2. Submit an XML sitemap to Google Search Console and Bing Webmaster Tools.
  3. Fix broken links and long redirect chains.
  4. Set canonical URLs so duplicate versions of a page do not compete.
  5. Write a unique title and meta description for every page.
  6. Keep pages fast and stable on mobile.

None of this is new, and that is the point. Technical problems that hurt classic SEO hurt AI visibility too, because the same pages have to be fetched, read and indexed first.

Where should you start?

Work in order: access first, then indexing, then content structure, then entity signals. There is no point polishing answers on pages that a crawler cannot reach. Then ask ChatGPT, Perplexity and Google a few of the questions your customers ask and note which sources they cite.

If you want a quick baseline, run a free Growli audit. It reads your site the way Google and AI crawlers do and checks robots.txt, AI crawler access (including firewall blocks), sitemaps, structured data, JavaScript-only content and more, with each issue rated High, Medium or Low. Nobody can promise a citation, but you can remove the reasons you would be skipped.

Frequently asked questions

Is AI SEO the same as GEO?

In practice, yes. GEO (generative engine optimization), AEO (answer engine optimization) and AI search optimization all describe making content easy for AI systems to find and cite. The terms are new and not used consistently, so focus on the work rather than the label.

Do I need special markup or an llms.txt file to appear in Google AI Overviews?

No. Google says there is no special schema and no AI-specific text file required for AI Overviews or AI Mode. A page needs to be indexed and eligible to show a snippet in normal Search. Structured data can still help search engines understand your page.

Will blocking GPTBot remove my site from ChatGPT search?

Not according to OpenAI. GPTBot is used for model training, while OAI-SearchBot is used for ChatGPT's search features, and each robots.txt setting is independent. To stay visible in ChatGPT search, keep OAI-SearchBot allowed.

How do I know if AI tools are citing my site?

Google includes AI Overviews and AI Mode traffic in the Search Console Performance report, but not as a separate filter. Bing Webmaster Tools has an AI Performance report for Copilot citations. You can also check referral traffic in analytics and test real questions in each assistant.

Does AI SEO replace traditional SEO?

No. AI search tools depend on crawling and indexing, which are classic SEO foundations. Think of AI SEO as classic SEO plus extra attention to AI crawler access, answer-first writing and consistent business facts.

Keep reading