GPTBot in robots.txt
GPTBot is run by OpenAI. Purpose: training. Source: OpenAI documentation.
What blocking GPTBot changes
Blocking it asks the operator not to use your pages for AI model training. Search and live-fetch agents, where the operator runs them, use separate tokens with their own rules.
Block GPTBot
User-agent: GPTBot Disallow: /
Block GPTBot from one folder
User-agent: GPTBot Disallow: /private/
Allow GPTBot
If no group names GPTBot, it follows your "User-agent: *" rules. To allow it while other bots are blocked:
User-agent: GPTBot Allow: /
How many sites block GPTBot?
Of the 49 SaaS sites in our State of AI readiness check on 2026-10-17, 3 block GPTBot from their homepage: figma.com, canva.com, loom.com.
More: the guide AI crawlers in robots.txt and the AI crawler checker.
Questions
What is GPTBot?
GPTBot is the user agent token for an agent run by OpenAI. Its purpose: training. See the OpenAI documentation linked on this page.
How do I block GPTBot?
Add a group to robots.txt with "User-agent: GPTBot" and "Disallow: /". The rule applies to agents that follow robots.txt.
How do I check if my site blocks GPTBot?
Run the free check on this site. It tests your robots.txt against 8 AI agents on your homepage. The $19 deep report tests 18 agents per path.
Get an email when AI crawlers, robots.txt rules, or llms.txt practice change. At most twice a month.
We send a confirmation link first. See the privacy policy.