OAI-SearchBot is OpenAI’s crawler for search functionality. Websites that want to appear as web sources in ChatGPT Search should therefore check that the bot is not accidentally blocked in robots.txt.
OAI-SearchBot and GPTBot are separate
OpenAI documents search crawling and model-training controls separately. A website can allow OAI-SearchBot for search while making a different decision about GPTBot.
What happens if OAI-SearchBot is blocked?
OpenAI states that sites excluding OAI-SearchBot are not surfaced as search results in ChatGPT Search, although navigational links may be handled separately.
How should robots.txt be configured?
If ChatGPT Search visibility is desired, OAI-SearchBot should be allowed to reach content that is intended to be crawlable. Training access should be documented as a separate business and legal decision.
Crawler access alone is not enough
A technically accessible page still needs to be relevant, understandable and credible as a source. Information architecture, expertise, entities and evidence remain critical.