Get cited by AI search.

Most of it is good SEO with sharper edges: let the right crawlers in, answer plainly, back claims with sources, and measure what gets cited.

A thick stack of folded newspapers
Photo by Tim Mossholder on Unsplashdithered by Cyborb

To get your site cited by ChatGPT and other AI search tools, first make sure each assistant’s search crawler can reach your pages. Then make those pages worth quoting: a direct answer near the top, specific facts with named sources, and one strong page per question. That is the core of generative engine optimization, and it is mostly good SEO with sharper edges.

Some popular advice has nothing behind it. Google says its AI features need no special files or markup, and in the one controlled study of these tactics, keyword stuffing did little or nothing. Here is how assistants choose sources, what the evidence supports, the real status of llms.txt, and how to track your citations.

The short version
  • Assistants cite pages they can find through search. Allow OAI-SearchBot, PerplexityBot and Claude-SearchBot, and stay indexable in Google and Bing.
  • Blocking a training crawler, such as GPTBot, does not remove you from search answers. Blocking a search crawler does.
  • In a KDD 2024 study, citing sources and adding quotes or statistics raised visibility in AI answers by 30 to 40%. Keyword stuffing did not help.
  • llms.txt is a proposal, used mostly for documentation. Google says its AI features do not need it.
  • Track citations with Bing’s AI Performance report, utm_source=chatgpt.com referrals and monthly spot checks.

How AI assistants choose which sources to cite

AI search works in two steps. First the assistant searches; Google’s AI features run several related searches at once, a technique Google calls query fan-out. Then the assistant reads the best passages and writes an answer that links to its sources. If your page is not in what the search step returns, nothing on the page matters.

Each company runs its own crawlers, and most separate search from model training:

AssistantAllow this to appear in answersTraining is controlled by
ChatGPT searchOAI-SearchBotGPTBot
PerplexityPerplexityBotNo training bot: Perplexity says PerplexityBot does not crawl for foundation models
ClaudeClaude-SearchBotClaudeBot
Google AI Overviews and AI ModeGooglebot, through normal Search indexingGoogle-Extended, which does not affect Search

OpenAI is explicit: sites that opt out of OAI-SearchBot “will not be shown in ChatGPT search answers”, though they can still appear as plain navigational links. Blocking GPTBot, by contrast, only signals that your content should not be used for training. Microsoft reports citations in Copilot and Bing’s AI summaries through Bing Webmaster Tools, covered below.

What gets picked also leans toward reference sources. In Pew’s March 2025 data, Wikipedia, YouTube and Reddit made up 15% of the sources in Google’s AI summaries, and government sites appeared three times as often as in regular results.

What the evidence says works

The best controlled evidence comes from the GEO study, which rewrote pages with different tactics and measured how much of each AI answer drew on them.

30 to 40%
Visibility gain in AI answers from citing sources, adding quotes or adding statistics
GEO study, KDD 2024
115%
Visibility gain for fifth-ranked pages that added citations
GEO study, KDD 2024
120%
More clicks per impression when a page is cited in a Google AI Overview
Seer Interactive, April 2026

Citing sources, adding quotations and adding statistics worked best. Making text more fluent and easier to understand added 15 to 30%. Keyword stuffing offered little to no improvement.

The gains were largest for pages that ranked lower. Adding citations lifted fifth-ranked pages by 115%, while top-ranked pages lost 30% on average. Results varied by topic, and the tests ran on a research benchmark, so read the numbers as direction, not a promise.

Two newer data points fill in the picture. Seer found that being cited in an AI Overview brings far more clicks than being left out. And in Ahrefs’ study of 75,000 brands, mentions on YouTube and across the web correlated with AI visibility much more than backlinks did, though correlation is not causation. For how AI answers changed clicks overall, see SEO in the age of AI.

Backed by evidence
  • Letting each assistant’s search crawler in, per their own docs
  • Citing sources and adding quotes and statistics (GEO study)
  • Plain, readable writing (GEO study)
  • One strong page per topic instead of near-duplicates (Bing)
Not proven or not needed
  • Keyword stuffing (little to no effect in the GEO study)
  • llms.txt as a citation lever (Google says AI features do not need it)
  • Special markup for AI answers (Google says none is needed)
  • Mass-produced AI pages (Google treats them as spam)

How to make a page worth citing

  1. Let the search crawlers in

    Check your robots.txt. Blocking a search crawler removes you from that assistant’s answers. The last rule below is optional: it keeps your pages out of OpenAI’s model training without affecting ChatGPT search.

    robots.txt
    User-agent: OAI-SearchBot
    Allow: /
    
    User-agent: PerplexityBot
    Allow: /
    
    User-agent: Claude-SearchBot
    Allow: /
    
    User-agent: GPTBot
    Disallow: /
  2. Answer the question in the first two sentences

    Lead with the answer in plain words, then explain. Readable text did better in the GEO study, and a clear opening sentence is easy for an assistant to lift into an answer.

  3. Back claims with sources, numbers and quotes

    Link to the sources you rely on. Give specific numbers with dates. Quote named experts where it helps. These were the three strongest tactics in the GEO study.

  4. Keep one strong page per question

    Merge near-duplicate pages and point the rest at the survivor with a canonical tag. Bing warns that AI systems group near-duplicate URLs and may pick an outdated version to represent the group.

  5. Tell search engines when you merge or remove pages

    Bing says duplicate and low-value URLs slow down how fast new information reaches AI summaries. It recommends IndexNow, a protocol for notifying search engines of changes, when you remove or consolidate pages.

  6. Earn mentions beyond your own site

    Brands that people talk about, on video and across the web, show up more in AI answers. Original data, useful tools and clear opinions are what get you mentioned.

What about schema markup and llms.txt?

Structured data, or schema, helps search engines understand a page and can earn rich results. For AI features, though, Google says you “don’t need to create new machine readable files, AI text files, or markup.” Use schema where it earns rich results, and do not expect it to win citations on its own.

llms.txt is a proposal, not a standard. Jeremy Howard suggested it in September 2024: a Markdown file at /llms.txt that gives language models a short, clean summary of a site, with links to key pages. A second version of the proposal followed in August 2026.

Its real use today is documentation. OpenAI, Anthropic and Google publish llms.txt files for their developer docs, and documentation platforms can generate one automatically. Chrome’s Lighthouse checks for the file in its agentic browsing audits, but the check is unscored, and a missing file counts as not applicable.

For AI search, the official guidance points elsewhere. Google says its AI features need no AI text files. OpenAI’s publisher FAQ asks you to allow OAI-SearchBot and does not mention llms.txt. Add one if developers read your docs through coding agents, since it costs little, but do not count on it for citations.

How to track whether AI search cites you

  • Bing Webmaster Tools. The AI Performance report, in public preview since February 10, 2026, shows how often your pages are cited in Copilot and Bing’s AI summaries, which pages, and the “grounding queries” the AI used to find them. It does not show clicks.

  • Analytics referrals. ChatGPT adds utm_source=chatgpt.com to the links it sends, so those visits are easy to separate in tools such as Google Analytics.

  • Google Search Console. Clicks from AI Overviews and AI Mode are counted under the “Web” search type, mixed in with regular results.

  • Server logs. Crawler visits show access, not citations, but a page no search bot has fetched cannot be cited.

Terminal
# Count visits from AI search crawlers and user-triggered fetchers
grep -Eio "OAI-SearchBot|ChatGPT-User|PerplexityBot|Perplexity-User|Claude-SearchBot|Claude-User" access.log | sort | uniq -c

Finally, check by hand. Run this prompt monthly in the assistants your customers use, and log which URLs appear.

PromptCitation check
I am researching [topic]. Answer this question: [the exact question your page answers].
Search the web and cite every source you use, with its full URL.

Answers change from run to run, so compare trends over months, not single replies. For sharper test prompts, see our prompt engineering guide. If you rely on these tools for your own research, our guide to AI research tools shows how to check every citation they give you.

FAQ

What is generative engine optimization?

It is the practice of shaping content so AI systems that write answers are more likely to use and cite it. The name comes from a 2023 research paper that measured which rewriting tactics raised a page’s visibility in AI answers.

Is GEO different from SEO?

Mostly not. Assistants search before they answer, so crawl access, indexing and quality still decide whether you are in the running. GEO adds a sharper focus on being quotable: clear answers, sourced facts and distinct pages.

Does llms.txt help me get cited by ChatGPT?

Not according to the published guidance. OpenAI’s publisher FAQ asks you to allow OAI-SearchBot and does not mention llms.txt, and Google says its AI features need no AI text files.

Should I block AI crawlers?

Block training crawlers if you do not want your content used to train models. At OpenAI and Google, that does not affect search answers. Blocking search crawlers, such as OAI-SearchBot or PerplexityBot, takes you out of those assistants’ answers.

Key takeaways
  • Get found first: let each assistant’s search crawler in and stay indexable.
  • Make pages quotable: answer first, cite sources, add numbers and quotes.
  • Keep one strong page per question, and tell search engines when you merge pages.
  • Treat llms.txt as a documentation convenience, not a citation lever.
  • Track citations in Bing, referrals in analytics, and answers by hand each month.

Next, read SEO in the age of AI, or build a website with AI that follows these rules from day one.

Sources
  1. GEO: Generative Engine Optimization, Aggarwal et al., KDD 2024
  2. Overview of OpenAI crawlers, OpenAI
  3. Publishers and developers FAQ, OpenAI Help Center, August 2026
  4. Perplexity crawlers, Perplexity
  5. Does Anthropic crawl data from the web?, Anthropic, April 2026
  6. AI features and your website, Google Search Central, December 2025
  7. Google’s common crawlers, Google, July 2026
  8. Google users are less likely to click on links when an AI summary appears, Pew Research Center, July 2025
  9. AIO impact on Google CTR: 2026 update, Seer Interactive, April 2026
  10. AI brand visibility correlations, Ahrefs, December 2025
  11. Does duplicate content hurt SEO and AI search visibility?, Microsoft Bing, December 2025
  12. Introducing AI Performance in Bing Webmaster Tools public preview, Microsoft Bing, February 2026
  13. The llms.txt proposal, llmstxt.org
  14. llms.txt audit, Chrome for Developers, May 2026
cyborb.ai

Stop reading about it. Build it.

Describe what you want in plain words. Cyborb plans the work, writes and runs the code, makes the assets, and puts the result online.

Download Cyborb

Free to start. No card required.