OpenAI’s dedicated crawler used to discover and surface websites in ChatGPT search results. Its permissions are independent of GPTBot.
OAI-SearchBot is OpenAI’s crawler used to help surface websites in ChatGPT search features. It is a search-discovery agent, distinct from GPTBot’s training-related role. Their access settings are independent, so a business should not treat one robots.txt choice as a universal policy for every OpenAI interaction.
Allowing the documented crawler can support eligibility for discovery, but it does not guarantee a citation, recommendation or visit. A working public page can still be absent from a particular answer. Access is one condition; the relevance of the request and the system’s selection process are different questions.
OpenAI’s documentation also distinguishes automatic search crawling from certain user-triggered visits. This is why a business should read the current agent descriptions rather than assume every request with an OpenAI-related identifier behaves the same way. Different controls can have different effects.
Takeaway: If ChatGPT search discovery is a goal, check the specific search crawler and infrastructure access. Keep that check separate from claims about actual visibility or training permissions.
Example
A supplier finds that its public compatibility guide is blocked by a broad bot rule. The team reviews whether OAI-SearchBot is unintentionally affected and makes the access decision deliberately. It then evaluates actual search observations rather than announcing that the change guarantees the guide will be cited.
What to check
Check robots.txt and relevant firewall or hosting rules using current provider documentation. Record the intended behavior before changing anything. A blocked search crawler is not a privacy mechanism for secret information, and an allowed crawler is not evidence that every answer contains the latest version of a page.
For more detail, see OpenAI crawler documentation.