ChatGPT-User
Operated by OpenAI · Powers ChatGPT
ChatGPT-User fetches a web page in direct response to a specific action a ChatGPT user takes — for example, asking the assistant to open a link, summarize a page, or check current information on a site. Unlike training crawlers, it acts on behalf of a real person, in real time.
What it does
When a person asks ChatGPT to look at a specific URL, summarize an article, or pull live details from a website, ChatGPT-User makes the request on their behalf. It only visits pages that are directly relevant to that one conversation — it does not crawl a site broadly or build an index.
Because the visit is tied to a real user need, blocking this agent typically breaks a legitimate feature for your visitors rather than protecting your content from anything.
Why we recommend keeping it open
- Blocking it can break a real user's ChatGPT experience on your site
- Not used to train OpenAI's models
- Behaves like a one-off visitor, not a bulk crawler
Identification & robots.txt
ChatGPT-User
Additional verification information is available in OpenAI's official documentation.
User-agent: ChatGPT-User
Allow: /