GPTBot - OpenAI's Web Crawler and How to Control Its Access

Categories: Open Source, OpenAI

In the ever-evolving world of artificial intelligence, data collection and usage have become central topics of discussion. OpenAI, a leading organization in the AI field, has introduced GPTBot, a web crawler that can access and analyze web content. What makes this initiative noteworthy is that OpenAI is letting website owners protect their data from being used to train its models. Here's a comprehensive look at GPTBot and how website owners can control its access.

What is GPTBot?

GPTBot is a web crawler developed by OpenAI. It's identified by the following user agent and string:

  • User agent token: GPTBot
  • Full user-agent string: Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; GPTBot/1.0; +OpenAI GPTBot)


GPTBot's primary function is to crawl web pages to potentially improve OpenAI's future AI models. The content accessed by GPTBot is filtered to exclude sources that require paywall access, gather personally identifiable information (PII), or contain text that violates OpenAI's policies.

While allowing GPTBot to access a site can contribute to the accuracy and safety of AI models, OpenAI has also provided options for website owners to disallow or customize GPTBot's access.

Disallowing GPTBot

Website owners who wish to block GPTBot from accessing their site can add the following lines to their site's robots.txt file:

User-agent: GPTBot
Disallow: /

Customizing GPTBot Access

To allow GPTBot to access specific parts of a site while blocking others, the following lines can be added to the robots.txt:

User-agent: GPTBot
Allow: /directory-1/
Disallow: /directory-2/

IP Egress Ranges

OpenAI's crawler makes calls to websites from a specific IP address block, as documented on OpenAI's website. This information may be useful for administrators monitoring or managing traffic from GPTBot.


GPTBot represents a thoughtful approach to data collection in AI development. By providing tools to control accessibility and protect data, OpenAI has acknowledged the importance of privacy and ethical considerations.

Whether to allow, restrict, or block GPTBot's access is a decision that rests with website owners. This flexibility reflects a growing trend towards transparency and control in the AI industry.

For more details about GPTBot and OpenAI's other initiatives, you can visit OpenAI's official website.