ChatGPT: what OpenAI documents, and what it does not.
OpenAI publishes how its crawlers work and how a site can be included in ChatGPT search. The pages cited here do not say how sources are chosen for an answer. This page keeps the two apart.
Appearing in ChatGPT means being used, and where possible cited, in an answer that ChatGPT gives when it searches the web. OpenAI documents one condition for inclusion in summaries and snippets: not blocking its search crawler. It gives no method for obtaining a mention.
Documentation last checked: October 4, 2026. It is re-checked every quarter, and each check is entered in the change log. These pages can change; the statements below are quoted or paraphrased from the sources linked, as they read on that date.
What is documented
| Statement | Source |
|---|---|
| "Any public website can appear in ChatGPT search." | OpenAI, publishers and developers FAQ |
| For content to be included in summaries and snippets, the site must not block OAI-SearchBot | OpenAI, publishers and developers FAQ |
| "Sites that are opted out of OAI-SearchBot will not be shown in ChatGPT search answers" | OpenAI, crawlers |
| If the URL of a disallowed page is obtained from a third-party search provider, OpenAI may surface just the link and the page title | OpenAI, publishers and developers FAQ |
| GPTBot crawls content that may be used in training | OpenAI, crawlers |
| ChatGPT-User acts on a user's request; because the action is user-initiated, robots.txt rules may not apply | OpenAI, crawlers |
ChatGPT adds utm_source=chatgpt.com to referral URLs |
OpenAI, publishers and developers FAQ |
A noindex tag can be used; the crawler must be allowed to crawl the page to read it |
OpenAI, publishers and developers FAQ |
What the cited sources do not cover
- How sources are selected or ordered for a given answer.
- Any way to request or guarantee a mention.
- How structured data is used, if it is.
- Whether, and when, a given answer uses web search.
Where the documentation is silent, this page says so and does not fill the gap.
Access and controls
OpenAI lists three user agents with different purposes: OAI-SearchBot for search and GPTBot for training, each set on its own line in robots.txt, and ChatGPT-User for actions a user triggers, for which OpenAI says robots.txt rules may not apply. Disallowing GPTBot is therefore a statement about training, not about search. A rule written to block "AI bots" in general can block the first without intending to. The general logic is set out in what Google and OpenAI document.
A test protocol for ChatGPT
This is a procedure to follow, not a result. No test of ChatGPT is reported on this page.
- Check access. Read robots.txt and confirm that
OAI-SearchBotis not disallowed. - Write the prompts. Entity prompts ("who is…"), category prompts and comparison prompts, in the user's words.
- Run each prompt several times, and note for each run whether the answer shows sources.
- Keep the full answer and the cited URLs, with date, language and location.
- Check referrals. In your analytics, look for visits carrying
utm_source=chatgpt.com. - Repeat at a fixed interval.
What to record for each run is listed in what is recorded; how many prompts a reading needs is in how many prompts make a reading. The rules a measurement must meet before it is published on this site are in the research methodology.
What to do with the result
- Not shown at all Check access first; then see not mentioned by ChatGPT.
- Shown, with wrong facts See inaccurate or outdated AI answers.
- Others are listed for your category See competitors recommended by AI.
- Referrals appear and you are not cited by name Read which pages receive them.
No one can guarantee a mention, a citation or a position in ChatGPT. What can be worked on is access, the clarity of the entity, the evidence on the page and the independent sources. The method is The Authority Architecture System.
Frequently asked questions
What does OpenAI say a site must do to appear in ChatGPT search?
Its FAQ for publishers states that any public website can appear, and that for content to be included in summaries and snippets the site must not block OAI-SearchBot. The pages cited here describe no way to guarantee a mention.
Does blocking GPTBot remove me from ChatGPT search?
Not according to OpenAI's crawler documentation, which separates GPTBot, for training, from OAI-SearchBot, for search. Each is set on its own line in robots.txt.
How do I know whether ChatGPT sends me visitors?
OpenAI states that ChatGPT adds utm_source=chatgpt.com to referral URLs. Those visits can be read in your analytics.
Other platforms
- Google AI Overviews Eligibility through Google Search, and the snippet controls.
- Google AI Mode Follow-up questions and query fan-out, with the same eligibility as Search.
- Gemini Google-Extended, a control separate from those of Google Search.
- Perplexity PerplexityBot, Perplexity-User and the recommendation to allow the first.
- Claude Three Anthropic crawlers with three different effects.
How the engines differ, and why each is measured separately, is set out in AI visibility. Related analyses are collected in Insights.
Understand how search and AI currently interpret your authority.
Request a confidential assessment of your visibility, entity signals, narrative consistency and source authority.
Request a Private AssessmentSelective engagements for leaders and organizations with material reputational stakes.