How to Get Cited by ChatGPT, Perplexity, Gemini and Copilot

To get cited by ChatGPT, Perplexity, Gemini and Copilot, first make sure each one can reach your site: allow OAI-SearchBot, PerplexityBot and Claude-SearchBot, and be well indexed in Google and Bing. Then publish clear, specific answers to buyer questions and keep facts about your company consistent across the web.

“How do we show up in ChatGPT?” is now one of the first questions B2B marketing leaders ask. The honest answer is that there is no single switch. Each assistant has its own way of finding sources, its own crawler or index, and its own settings. The good news is that the work overlaps heavily, and most of it is the same work that improves search visibility in general.

How do AI assistants find the sources they cite?

When an assistant searches the web, it retrieves pages from a search index, either its own or a partner’s, and grounds its answer in them. Being in that index is the entry ticket.

Google’s own AI Overviews and AI Mode follow Google Search rules instead, which we cover separately in how to get cited in Google AI Overviews.

How does ChatGPT search choose sources?

ChatGPT search surfaces websites found by OAI-SearchBot. If you block that crawler, OpenAI says your site will not be shown in ChatGPT search answers, though it can still appear as a navigational link.

OpenAI documents three crawlers that matter here. OAI-SearchBot is for search. GPTBot crawls content that may be used to train OpenAI’s foundation models. ChatGPT-User handles actions a user asks for, such as visiting a page during a conversation; because those requests are user-initiated, OpenAI says robots.txt rules may not apply to it. The settings are independent: you can allow OAI-SearchBot to appear in search while disallowing GPTBot for training.

OpenAI also recommends allowing requests from its published IP ranges. That matters for B2B sites behind a web application firewall or a bot-protection service, which can block legitimate crawlers even when robots.txt allows them.

How does Perplexity find and cite pages?

Perplexity cites pages from its own index, built by PerplexityBot. Perplexity says PerplexityBot is not used to train AI foundation models.

Perplexity recommends allowing PerplexityBot and its published IP ranges if you want to appear in its results, and provides instructions for common firewalls. Its second agent, Perplexity-User, visits pages when a user’s question requires it and, because a person requested the fetch, generally ignores robots.txt. Perplexity says changes to robots.txt can take up to 24 hours to be reflected.

How does Gemini use your content?

The Gemini app grounds many answers with Google Search. Google’s Google-Extended token controls whether your content may be used for that grounding and for training Gemini models.

Google-Extended has no crawler of its own; it is a robots.txt token used in a control capacity. Google says it affects training of Gemini models and grounding in Gemini Apps and Grounding with Google Search on Vertex AI, and that it has no effect on inclusion or ranking in Google Search. That creates a real trade-off: blocking Google-Extended to opt out of training may also reduce how your content is used in Gemini app answers. Decide deliberately rather than copying a robots.txt template.

How does Microsoft Copilot pick sources?

Copilot relies on Bing. Strong Bing indexing, fresh content and clear pages are the foundation, and Bing Webmaster Tools now shows how often you are cited.

Bing’s Webmaster Guidelines say the fundamentals that help pages perform in Bing search also apply to how content is selected for AI answers. Bing recommends surfacing key information early, keeping each URL focused on one topic, writing content that stands on its own and can be verified, and defining entities consistently. It also points to IndexNow, a protocol that notifies participating search engines when pages are added, updated or removed, so AI answers reference the current version.

Bing’s AI Performance report, in public preview since February 2026, shows total citations, average cited pages, the grounding queries used to retrieve your content and page-level citation counts. In June 2026 Microsoft added Intents, Topics, Citation Share and Compare views in preview.

What about Claude?

Anthropic separates three agents: ClaudeBot for training data, Claude-SearchBot for search quality and Claude-User for user-requested fetches. All honor robots.txt.

Anthropic says disabling Claude-SearchBot prevents your content from being indexed for search, and disabling Claude-User prevents retrieval for user questions, either of which may reduce visibility. Blocking only ClaudeBot opts your future content out of training without affecting the other two. Anthropic notes that you need to set rules for every subdomain you want to cover.

What works across every assistant?

Crawlable pages, clear answers, consistent facts and genuine third-party corroboration. These matter for all engines at once.

1. Answer real questions directly. Put a concise answer at the top of the section, then the detail. See answer-first content.
2. Be specific. Name the industries, company sizes, integrations and regions you serve. Vague claims give an assistant nothing to quote.
3. Keep facts consistent. Your company description, services and locations should match on your site, LinkedIn, directories and partner pages. See entity SEO for AEO.
4. Earn real mentions. Being discussed in industry publications, partner pages and communities helps assistants corroborate you. Fake or paid mentions disguised as editorial do not, and Google warns they are not as helpful as they seem.
5. Keep content fresh. Update pages when facts change, and use IndexNow for Bing and its partners.
6. Keep pages fast and readable without scripts. Important text should be in the HTML, not only rendered later by JavaScript.

What can’t you control?

You cannot control which sources an assistant picks for a given answer, and the same question can produce different answers on different days.

Answers vary by wording, user context and model updates. Treat single screenshots with caution and look at trends over many questions. Also avoid trying to steer assistants with hidden text or instructions addressed to AI; Bing’s guidelines treat this kind of prompt injection as a violation. If a vendor guarantees organic placement in ChatGPT or Perplexity, ask how, then walk away.

Where should a B2B team start?

Start by checking your robots.txt and firewall for the crawlers in the table above, then pick 20 buyer questions and see who each assistant cites today. Our AI crawler guide includes a sample robots.txt, and our GEO services and AEO services pages explain how we run this for clients across the United States, Canada and India. Or start with a free audit.

Frequently asked questions

Does blocking GPTBot remove my site from ChatGPT search?

No. OpenAI treats them separately. GPTBot relates to training its foundation models, while OAI-SearchBot is used to surface websites in ChatGPT search. You can disallow GPTBot and still allow OAI-SearchBot.

How long does it take ChatGPT or Perplexity to respect a robots.txt change?

Both OpenAI and Perplexity say it can take about 24 hours for their systems to adjust after you update robots.txt.

Is Microsoft Copilot based on Bing?

Copilot’s web answers draw on Bing’s index, and Bing Webmaster Tools reports citations across Copilot and AI summaries in Bing. Being well indexed in Bing, and keeping it updated with IndexNow, is the foundation for Copilot visibility.

Can I pay to be cited in ChatGPT or Perplexity answers?

Organic citations cannot be bought. Some platforms are introducing labeled advertising, such as ads in ChatGPT, but those are separate from the sources an answer cites. Be wary of anyone selling guaranteed organic AI placement.

Why does ChatGPT describe my company incorrectly?

AI answers repeat what they find. Outdated pages, inconsistent descriptions or old third-party profiles are common causes. Fix the facts on your own site first, keep important pages crawlable, then update the profiles and directories you control.

Sources

• OpenAI: Overview of OpenAI crawlers
• Perplexity: Perplexity crawlers
• Anthropic: Does Anthropic crawl data from the web, and how can site owners block the crawler?
• Google: Common crawlers, including Google-Extended
• Bing Webmaster Guidelines
• Bing Webmaster Blog: Introducing AI Performance (February 2026)
• Bing Search Blog: Intents, Topics, Citation Share and Compare (June 2026)
• IndexNow