Question

Should I allow GPTBot?

Updated June 7, 2026

GPTBot is OpenAI's model-training crawler, not the one that powers ChatGPT search. Blocking GPTBot does not reduce your ChatGPT search visibility — that depends on OAI-SearchBot and ChatGPT-User. Allow it only if you are comfortable with your content being used to train future models.

What each OpenAI crawler does

• OAI-SearchBot: indexes pages so they can surface and be cited in ChatGPT search. Allow this for AI answer visibility.
• ChatGPT-User: fetches a page when a user (or a GPT) clicks or asks about it in real time. Allow this so live lookups work.
• GPTBot: collects content to train future OpenAI models. Independent of search visibility.

Decision in one line

• Want AI search visibility, neutral on training: allow OAI-SearchBot and ChatGPT-User, block GPTBot.
• Happy to contribute to training too: allow all three.
• Strict content control: block GPTBot, keep search bots allowed so you stay quotable.

robots.txt: visibility without training

User-agent: OAI-SearchBot
Allow: /

User-agent: ChatGPT-User
Allow: /

User-agent: GPTBot
Disallow: /

Verify the policy is read correctly

curl -s https://yourdomain.com/robots.txt | grep -A1 -Ei 'gptbot|oai-searchbot|chatgpt-user'

# Then confirm real crawler behavior in your access log
grep -Ei 'GPTBot|OAI-SearchBot' /var/log/your-access.log | awk '{print $1, $7}' | sort | uniq -c | sort -nr | head

Common mistake

Teams block GPTBot to "keep AI out" and then wonder why ChatGPT never cites them. Search visibility and training are separate decisions controlled by separate tokens — blocking the training bot does not remove you from ChatGPT answers, and blocking the search bot does.

Related pages: should I allow OAI-SearchBot, OpenAI crawler guide, why GPTBot is still blocked.

Build your AI robots.txt