To get your website cited by ChatGPT, let OpenAI's search crawler, OAI-SearchBot, reach it through your robots.txt file and through OpenAI's published IP ranges at your host or content delivery network (CDN), keep your pages indexable and well organised, and publish original content worth citing. OpenAI's help page is plain that none of these steps guarantees placement.
Karim Kamal, our Founder & CEO, describes AI assistants like ChatGPT as already "answering our questions and shaping our decisions before we ever reach a website." At Amplifyr, we built Echo, our agent, around that change. Echo runs in the background, surfaces the gaps costing you visibility, such as a page models can't retrieve, and produces the output end to end, ready for your team to review.
Start with crawler access, not content tweaks
Can ChatGPT cite a page its search crawler is not allowed to read? Not in its search answers. Access comes before everything else.
OpenAI's crawler overview lists four agents: OAI-SearchBot for ChatGPT search, GPTBot for training data, ChatGPT-User for fetches a user starts, and OAI-AdsBot. OpenAI says its OAI-SearchBot and GPTBot settings are each "independent of the others." Sites that opt out of OAI-SearchBot will not be shown in ChatGPT search answers, though OpenAI notes they can still appear as navigational links. No amount of content work makes up for a blocked search crawler.
Because the settings are independent, you can allow OAI-SearchBot for search and disallow GPTBot to signal that your content should not be used to train OpenAI's foundation models. ChatGPT-User works differently. A person triggers those visits, robots.txt rules may not apply, and OpenAI says the agent is not used to decide what can appear in Search. A ChatGPT-User hit in your server logs tells you nothing about search eligibility.
A starting robots.txt block for search access without training consent looks like this:
`
User-agent: OAI-SearchBotAllow: /User-agent: GPTBotDisallow: /
`
Check it against any more specific rules already in your file before you rely on it.
Robots.txt is only half of access. OpenAI's Help Center asks site owners to confirm that their host or CDN allows traffic from OpenAI's published searchbot IP addresses. If a firewall or CDN rule drops those requests, the crawler never reads your robots.txt at all.
Crawler | Purpose | What blocking it does | Where to configure |
OAI-SearchBot | ChatGPT search | Site is not shown in ChatGPT search answers | robots.txt + host/CDN IP rules |
GPTBot | Training data collection | Signals content should not train OpenAI's models | robots.txt |
ChatGPT-User | Fetches a user starts | Rules may not apply; OpenAI does not use it to decide what appears in Search | robots.txt (may not apply) |
Other assistants publish similar controls. Anthropic's crawler page says blocking Claude-SearchBot or Claude-User may reduce your site's visibility in Claude's search results and user-directed web search.
How ChatGPT chooses sources once your page is reachable
Access gets a page considered. It does not get it cited. According to OpenAI's Help Center, ChatGPT "ranks search results using multiple factors" meant to help users find relevant, reliable information.
To answer a question, ChatGPT search "sometimes partners with other search providers." When it does, it typically rewrites the question into one or more targeted queries and sends them to those providers.
So what decides whether your page is the one picked? OpenAI does not publish a citation formula, and neither do Google or Microsoft. Those two do publish what makes a page eligible for their own AI answers, and what kind of content they want.
Make pages eligible in Google and Bing
Google's guide to generative AI features says a page must be indexed and eligible to be shown in Google Search with a snippet. Search Console's Search generative AI control, rolled out to all websites worldwide as of 31 August 2026, includes every property by default. There is nothing to do unless someone switched your property, or a parent property it inherits from, to exclude, which removes traffic and impressions from these features.
Microsoft's Bing Webmaster Guidelines cover Bing search, Copilot and grounding API results. They are Microsoft's guidance for its own systems, not a statement about ChatGPT. Bing asks for XML sitemaps with freshness signals such as lastmod and for crawlable internal links. It also recommends IndexNow notifications when content changes. Its guidelines state that generative engine optimisation (GEO) "does not guarantee grounding or citations" in AI experiences.
Where should you start if a page never gets cited? For ChatGPT, confirm OAI-SearchBot is allowed in robots.txt and at your CDN. For Google's AI features, confirm the page is indexed with a snippet and that the Search Console control is set to include. For Bing and Copilot, make sure your sitemap's lastmod values move when the content does.
Publish the kind of material models can quote
Access and indexing put a page in the candidate pool. From there, Google says original, useful content will likely do more for your presence in its AI search, over time, than any other step in its guide.
Original, first-hand content. Google's guide asks for "a unique point of view" and warns against recycling what others have already said or what a generative AI model could easily produce. Bing asks for original, authoritative content. Its guidelines add that URLs are more likely to be selected for grounding and citations when content stands on its own.
Clear structure. Google notes that people appreciate pages organised by paragraphs and sections, with headings that give a clear structure. It also says there is no requirement to break content into tiny pieces for AI to understand it. Bing asks you to place essential information near the top of the page and to avoid long introductions before the main topic.
Fan-out, handled honestly. Google defines query fan-out as a set of concurrent, related queries the model generates to fetch more results for the user's question. Answering the real sub-questions your buyers ask helps. Building pages for every variation primarily to manipulate rankings or AI responses breaches Google's scaled content abuse spam policy.
Real mentions. Google's generative AI features can show what is said about products and services in blogs, videos and forum discussions. Seeking inauthentic mentions, Google adds, "isn't as helpful as it might seem." Bing rules out link schemes and artificial promotion.
Freshness. Google describes retrieval-augmented generation (RAG) as relying on its core Search ranking systems to retrieve relevant, up-to-date web pages. Bing asks you to update content when facts or guidance change and to notify IndexNow. Neither names a time window.
Local and language signals. Google's guide says Google Business Profiles can help your products and services show in AI responses. If you run Arabic and English versions of a page, Google's guidance on localized versions asks you to tell Google about each variation so Search can point users to the right one.
John Mueller of Google Search wrote on 15 May 2026 that Google published its guide to help website owners, SEOs and developers "understand how to optimize their content for appearance in generative AI features in Search". It sets out eligibility rules and content principles. It does not promise citations.
What structured data and llms.txt actually do
Do you need schema or an llms.txt file to be cited? For Google Search, Google's answer is no. Its guide says you don't need new machine-readable files, AI text files, special markup or Markdown, because Google Search itself doesn't use them.
Structured data "isn't required for generative AI search", in Google's words, though Google still recommends it for rich results, and it should match the visible text on the page. Organization markup can help Google "disambiguate your organization" in search results, which matters for brands with common names. Bing says structured data "does not guarantee visibility." OpenAI's crawler and help pages give no guidance on llms.txt or schema, so Google's position covers Google Search only.
Measure citations with a repeatable test
How do you know whether any of this worked? A single prompt result is an anecdote. Write down the questions your buyers ask and run that same set at regular intervals with the same model settings. Log which pages each answer cites, and count how many runs cite you out of the total. In Google, Search Console's generative AI performance report shows impressions in these features by page, country, device and date.
This is the work we built Amplifyr for. Echo scans every prompt, model and source that matters to your category and surfaces the gaps, such as a question AI answers without you. Each signal comes with a next step ranked by impact, and Echo then produces the output for your team to review. Technical Optimization handles crawlability and schema fixes, while competitive AI intelligence shows which sources AI trusts in your category.
Book a demo to see how Echo finds your biggest AI search gap.
Frequently asked questions
Can I block AI training and still appear in ChatGPT search?
Yes. OpenAI treats GPTBot, its training crawler, and OAI-SearchBot, its search crawler, as separate settings. One robots.txt file can disallow the first and allow the second.
How long does a robots.txt change take to reach ChatGPT search?
OpenAI says it can take about 24 hours from a robots.txt update for its systems to adjust. If your CDN or firewall still blocks OpenAI's searchbot IP addresses, the change makes no difference.
Do I need an llms.txt file?
Not for Google Search, which Google says does not use AI text files. OpenAI's crawler and help pages give no guidance on it. Accessible, well-structured pages come first.
Does schema markup get a page cited?
No platform says it does. Google calls it optional for generative AI features, and Bing says it does not guarantee visibility. Use Organization markup to help Google identify your brand, and keep it matched to what the page shows.
References
OpenAI. "Overview of OpenAI Crawlers." https://developers.openai.com/api/docs/bots
OpenAI Help Center. "Searching the web with ChatGPT." https://help.openai.com/en/articles/9237897-searching-the-web-with-chatgpt
Google Search Central. "Google's Guide to Optimizing for Generative AI Features on Google Search." https://developers.google.com/search/docs/fundamentals/ai-optimization-guide
Google Search Central. "AI Features and Your Website." https://developers.google.com/search/docs/appearance/ai-features
Google Search Central Blog. "A new resource for optimizing for generative AI in Google Search." 15 May 2026. https://developers.google.com/search/blog/2026/05/a-new-resource-for-optimizing
Search Console Help. "Search generative AI control." https://support.google.com/webmasters/answer/16908024?hl=en
Search Console Help. "Generative AI performance report (Search)." https://support.google.com/webmasters/answer/16984139?hl=en
Google Search Central. "Organization Schema Markup." https://developers.google.com/search/docs/appearance/structured-data/organization
Google Search Central. "Localized Versions of your Pages." https://developers.google.com/search/docs/specialty/international/localized-versions
Microsoft Bing. "Webmaster Guidelines." https://www.bing.com/webmasters/help/webmaster-guidelines-30fba23a
Microsoft Bing. "Why IndexNow." https://www.bing.com/indexnow
Anthropic. "Does Anthropic crawl data from the web, and how can site owners block the crawler?" https://support.claude.com/en/articles/8896518-does-anthropic-crawl-data-from-the-web-and-how-can-site-owners-block-the-crawler
Share