In short. ChatGPT-User identifies certain user-initiated web requests, for which OpenAI says robots.txt rules may not apply.

Definition: As of 17 October 2026, OpenAI documents ChatGPT-User as a user-agent, an identifier sent with web requests, for certain user-initiated actions in ChatGPT and custom GPTs. A question can prompt a visit to a web page using this identifier.

Key facts

OpenAI's crawler documentation assigns ChatGPT-User to user-triggered activity and excludes automatic web crawling from its role.
The same documentation says robots.txt rules may not apply because a user initiates these actions.
OpenAI directs search opt-outs and automatic crawling controls to OAI-SearchBot.
The Robots Exclusion Protocol standard describes robots.txt rules as separate from access authorization.

How is a ChatGPT-User request identified?

OpenAI publishes this full user-agent string:

Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko); compatible; ChatGPT-User/1.0; +https://openai.com/bot

For a website log review, ChatGPT-User/1.0 is the identifying substring to look for. OpenAI also links its published IP addresses from the crawler documentation, providing a separate reference for comparing a request's source address.

The standard's user-agent section explains that crawlers choose a name called a product token. That token should occur within the identification string sent to the website. The whole browser-style string and the shorter bot name therefore serve different purposes when reviewing requests and writing crawler rules.

How does it differ from OpenAI's crawlers?

OpenAI's crawler overview assigns search to OAI-SearchBot and potential training collection to GPTBot. Their robots.txt settings are independent. ChatGPT-User has a separate role for user actions; OpenAI excludes it from determining search inclusion.

What control does robots.txt provide?

A robots.txt file expresses crawling rules. RFC 9309 organizes them into groups: user-agent lines followed by rules describing which paths a crawler may access. Its access-method section requires the file at the service's top-level /robots.txt path, encoded as UTF-8 and served as text/plain.

For ChatGPT-User, OpenAI's user-initiated-request qualification matters when interpreting such rules. A robots.txt entry alone is insufficient protection for private content. The standard's security section recommends an access-control measure such as HTTP authentication. It also warns that listing paths in robots.txt exposes those paths publicly.

FAQ

Can I block ChatGPT-User with robots.txt?

OpenAI's documentation says robots.txt rules may not apply to these user-initiated requests. For private pages, the protocol standard recommends an access-control measure such as authentication.

Which bot controls ChatGPT search opt-outs?

OpenAI identifies OAI-SearchBot as the robots.txt agent for search opt-outs and automatic crawling.

Sources

Check how a page reads to answer engines. Open the GEO score analyzer