GEO · How-to

How to Get Cited by ChatGPT (and Perplexity) in 2026

How ChatGPT, Perplexity and Copilot pick the sources they cite, which AI crawlers to allow, and the steps to become the page they quote — built from how answer engines actually work.

By Christopher TaylorFounder, Black & Gold SEOLast updated 9 min read

The short answer

To get cited by ChatGPT, become the kind of source it already trusts: an authoritative, well-referenced page on a clearly-defined entity, written answer-first so the model can lift a clean, attributable passage. Allow the AI search crawlers (OAI-SearchBot, GPTBot) to access your pages, keep your author and brand entities consistent, and make each page trivially quotable. Perplexity adds a strong recency and community bias on top of the same fundamentals.

Key takeaways

  • ChatGPT cites sources when it searches/browses — so being in its live results is the prerequisite.
  • Allow OAI-SearchBot and GPTBot in robots.txt, or you're invisible to ChatGPT search.
  • Authority + clear entities + answer-first passages are what get selected as citations.
  • Perplexity rewards freshness and community presence; Copilot leans on the Bing web.

Being named inside a ChatGPT or Perplexity answer is becoming one of the most valuable placements in search — the visitor arrives already having seen you recommended. This is the engine-specific companion to our GEO guide.

Does ChatGPT cite sources, and how does it choose?

When ChatGPT answers from its search capability, it retrieves live pages and cites the ones it used. The model favors sources that are established, well-referenced and clearly about a single topic — the same authority signals that win classic search, applied to passage selection. So your job is two-fold: be retrievable by its crawler, and be the cleanest, most-quotable answer once retrieved.

Step 1: Let the AI crawlers in

You can’t be cited if the engine can’t read you. Confirm your robots.txt allows the AI search crawlers — and note that search crawling is governed separately from training use, so you can allow one without the other.

Crawler (user-agent)OperatorPurpose
OAI-SearchBotOpenAISurfacing pages in ChatGPT search results
GPTBotOpenAICrawling (can be used for training)
PerplexityBotPerplexityIndexing pages for Perplexity answers
Google-ExtendedGoogleControls Gemini/Vertex AI training use

Step 2: Become the page worth quoting

  • Answer first, in a liftable chunk. Give a direct, ~40-word answer near the top that stands on its own when quoted out of context.
  • Show real expertise. Named author with genuine credentials, first-hand detail, and a clear About/editorial footprint. See our explainer on AEO for the structural side.
  • Cite primary sources. A visible chain of references signals trust to both readers and models — and the GEO study found citations and quotations lifted AI inclusion.
  • Keep entities consistent. Same author name, photo and profile links across the web so the engine can resolve and trust you as one entity.
  • Consider an llms.txt. A curated map of your best content is a low-cost AI-readiness signal — see our llms.txt guide.

ChatGPT vs Perplexity vs Copilot: what differs?

EngineSource biasWhat to emphasize
ChatGPT SearchEstablished, well-referenced pagesAuthoritative cornerstone content, entity consistency
PerplexityRecency + community (e.g. Reddit)Freshness, clear sourcing, community presence
Microsoft CopilotThe Bing indexSolid Bing-eligible SEO + extractable answers

Sources & further reading

Keep reading

Questions

Frequently asked

Does ChatGPT actually cite sources?

Yes. When ChatGPT uses its search/browse capability it returns answers with linked citations to the pages it drew from. Answers generated purely from training data may not cite, which is why being in its live search results matters.

Do I have to let GPTBot crawl my site?

If you want to be eligible for citation in ChatGPT's search results, yes — allow the relevant crawlers (OAI-SearchBot for search, GPTBot) in robots.txt. Blocking them removes you from that surface. Allowing them is separate from allowing training use.

What sources does ChatGPT prefer?

It leans toward established, well-referenced pages with clear authority and unambiguous entities — the same things that earn trust in classic search. Thin or anonymous pages are less likely to be selected as a citation.

How is getting cited by Perplexity different?

Perplexity weights recency and community sources (like Reddit) more heavily, and surfaces citations very visibly. Fresh, clearly-sourced content and a real community presence help more there than on ChatGPT.

Win the answer, not just the link

Black & Gold SEO finds, writes, and applies the on-page and entity fixes that get you cited in AI answers and ranked in classic search — evidence-grounded, and shipped to your site via one snippet.