What GPTBot is
GPTBot is the crawler OpenAI uses to gather publicly available web content that may be used to train its foundation models. It identifies itself with a user-agent string containing GPTBot and respects robots.txt.
OpenAI runs separate agents for other jobs. OAI-SearchBot crawls pages so they can be shown and cited in ChatGPT search results, and ChatGPT-User fetches a page when a user asks ChatGPT to visit it.
Should you block GPTBot?
Blocking GPTBot keeps your content out of future training data but does not remove you from ChatGPT search. Many publishers block GPTBot while allowing OAI-SearchBot, so they can still be cited and earn referral traffic.
If AI visibility matters to your business, check that you have not blocked the search agents by accident with a broad User-agent: * rule or a firewall bot filter.
How to block or allow it
Add a group for each agent to robots.txt. A specific group overrides the * group, so User-agent: GPTBot with Disallow: / blocks training while the rest of your rules stay unchanged. Changes take effect the next time the crawler fetches robots.txt, usually within about a day.
Other AI crawlers work the same way: ClaudeBot for Anthropic, PerplexityBot for Perplexity, Google-Extended to opt out of Gemini training without affecting Google Search, and CCBot for Common Crawl.
Example
Block training, allow ChatGPT search:
User-agent: GPTBot
Disallow: /
User-agent: OAI-SearchBot
Allow: /
User-agent: ChatGPT-User
Allow: /