Key point. GPTBot is OpenAI's AI training crawler. It governs use of your content for model training. It follows robots.txt.
GPTBot at a glance
These are the facts our registry holds for GPTBot, checked against OpenAI's documentation on October 14, 2026.
| Vendor | OpenAI |
| Purpose | AI training |
| Type | User agent |
| Follows robots.txt | Yes; source: community ai.robots.txt list (commit 9ad8a47e) |
| Vendor documentation | OpenAI crawler page, checked October 14, 2026 |
| Published IP ranges | openai.com/gptbot.json |
What is GPTBot?
GPTBot is OpenAI's AI training crawler. It governs use of your content for model training.
In OpenAI's own words (OpenAI's crawler documentation, read October 14, 2026):
- “It is used to crawl content that may be used in training our generative AI foundation models.”
OpenAI uses other names for other jobs: OAI-SearchBot (AI search); ChatGPT-User (AI user fetch); OAI-AdsBot (other). Each takes its own robots.txt group.
Does GPTBot follow robots.txt?
Yes, GPTBot follows robots.txt: our registry records this from the community ai.robots.txt list (commit 9ad8a47e).
What OpenAI says (OpenAI's crawler documentation, read October 14, 2026):
- “Disallowing GPTBot indicates a site’s content should not be used in training generative AI foundation models.”
A robots.txt rule records your request. It does not prove a crawler honours it, and firewall or CDN bot settings can refuse a crawler that robots.txt allows, so check those too.
How do I block GPTBot in robots.txt?
Add a group that names GPTBot and disallows the whole site:
User-agent: GPTBot
Disallow: /To keep it out of one folder only, disallow that path instead, for example Disallow: /private/.
How do I allow GPTBot?
GPTBot is allowed by default: if no group names it and no User-agent: * rule blocks it, it may read every page. A crawler obeys the most specific group that names it and falls back to User-agent: * only when no group names it (RFC 9309), so if a catch-all rule blocks crawlers, add a group for it:
User-agent: GPTBot
Allow: /How can I tell if a request really comes from GPTBot?
Match the request's IP address against OpenAI's published list, because user agents can be faked: https://openai.com/gptbot.json.
How often do startup websites block GPTBot?
Our AI Crawler Access Report read the robots.txt files of 600 sampled startup websites on October 7, 2026. This is its row for GPTBot, across the 569 sites whose answer it could read:
| Crawler | Type | Blocked on at least one path | Blocked on all six paths | Named in the file |
|---|---|---|---|---|
| GPTBot | AI training | 11 of 569 sites (1.9%; 95% interval 1.1% to 3.4%) | 9 of 569 sites (1.6%; 95% interval 0.8% to 3.0%) | 98 of 569 sites (17%; 95% interval 14% to 21%) |
“Blocked” means the file disallows the crawler on at least one of six common paths (/, /pricing, /blog/, /docs/, /about, /product).
The report's headline figure, for the three AI search crawlers it tracked: “3 of 569 startup websites (0.5%; 95% interval 0.2% to 1.5%) block at least one AI search crawler (Elementera AI Crawler Access Report, October 2026).”
The figures describe that sample of Y Combinator-backed software companies on the day we fetched it. The method, the intervals and the full table are in the AI Crawler Access Report.
What the AI Crawler Access Report adds about GPTBot:
No. OpenAI documents GPTBot and OAI-SearchBot as separate settings: GPTBot covers training, and OAI-SearchBot decides whether the site can appear in ChatGPT search answers. You can block GPTBot and allow OAI-SearchBot. In our sample, 15 of 569 sites (2.6%; 95% interval 1.6% to 4.3%) block training while leaving every AI search crawler open, and none do the reverse.
How do I check whether my robots.txt blocks GPTBot?
Paste your robots.txt into our free checker and it shows, crawler by crawler, whether GPTBot may read the path you choose. Nothing you paste leaves the page.