AI SearchOpenAI
Last verified: 2026-07-16 · maintained by Unsourced
ChatGPT-User is the fetch a ChatGPT user triggers when the assistant browses to a page on their behalf. It's a live, user-initiated request that can cite and link your page.
ChatGPT-User isn't a background crawler at all — it fires only when a person inside ChatGPT asks something that makes the assistant open a specific URL for them. Its traffic tracks real user intent rather than a crawl schedule, and the page it reads can be quoted and linked straight back into that conversation. Because each hit is a human-initiated visit, blanket-blocking it can cut you out of answers your own audience is actively requesting.
ChatGPT-User can be checked two independent ways, and a real request passes both. Its source IP has to fall inside the 266 ranges OpenAI publishes, and it has to forward-confirm by reverse DNS: the IP resolves to openai.com, and that hostname resolves back to the same IP. Anything carrying the ChatGPT-User user-agent that fails either test is an impostor, whatever the header says.
Recommended: bait. Every request is a real ChatGPT user browsing on their own behalf, so blocking it cuts you out of answers your own audience is actively asking for — not out of any training corpus.
If you do choose to act in robots.txt (which crawlers honour but don't enforce):
# ChatGPT-User: recommended to ALLOW — blocking can cost you AI visibility User-agent: ChatGPT-User Disallow:
Is ChatGPT-User a crawler or a real person's visit?
Neither exactly — it's the fetch ChatGPT performs when a user asks it to open a specific page. Each hit follows a real person's request rather than a crawl schedule, so the traffic maps to genuine intent.
Why does ChatGPT-User ignore my crawl-delay?
Because it isn't bulk-crawling — a person triggered it live, so rate rules written for schedulers don't apply. You can still confirm it's genuine via OpenAI's published IP ranges or reverse DNS to openai.com.
Will blocking ChatGPT-User cost me anything?
It can. Blocking it removes your pages from answers your own visitors are actively asking ChatGPT to fetch, so you lose live, intent-driven exposure rather than just a background crawl.
Unsourced checks each crawler against published ranges andreverse DNS, and shows where AI search cites you instead of a competitor.
Check your site free →10-day free trial · no card required · cancel anytime