ChatGPT
ChatGPT is the conversational assistant made by OpenAI. When it answers with search switched on, it runs a query, fetches a handful of pages and writes an answer that names sources, so a business can be recommended, cited or left out depending on what its site served a moment earlier.
How ChatGPT reaches a website
OpenAI publishes three product tokens, and they do different jobs. OAI-SearchBot indexes pages so they can be surfaced in ChatGPT search. GPTBot collects material that may be used to train models. ChatGPT-User fetches a page when a person in the conversation asks for it. Each is documented in OpenAI’s crawler documentation, which also says that disallowing OAI-SearchBot removes a site from ChatGPT search answers and that ChatGPT-User is not used to decide search inclusion.
The consequence is easy to get backwards. A robots.txt line that blocks GPTBot keeps a site out of training and changes nothing about whether ChatGPT can cite it; a line that blocks OAI-SearchBot does the reverse. The matching rules are those of RFC 9309: groups by product token, the most specific path winning, allow taking a tie.
What this site has and has not measured
This site has not run its question panel against ChatGPT at scale. Its one-question probe of 9,240 businesses was put to Claude, and the 14.9% who were named is a Claude number; it is reported in the 9,483-site audit and does not transfer to ChatGPT. What does transfer is the readiness data from the same audit, because a fetch is a fetch: 61.5% of the sites had no question-and-answer content for any assistant to lift, and 7.2% sat behind bot protection that served a challenge page to the crawler.
A question people ask
Does blocking GPTBot keep a business out of ChatGPT search answers?
No. GPTBot is the training crawler. ChatGPT search relies on OAI-SearchBot, and OpenAI states that a site which disallows that token will not appear in its search answers. To stay citable while opting out of training, allow OAI-SearchBot and disallow GPTBot, each in its own robots.txt group.
Related terms
Sources
Crawler tokens and their jobs: OpenAI’s crawler documentation. Matching rules: RFC 9309. Audit figures are this site’s own, computed from its audit records; the probe used one assistant, one question per business, and its percentages describe the sample.