Measured 2026-07-09 · 3 of 10 K-beauty export brands block ChatGPT's crawler · 2 are invisible to AI · every number reproducible
HomeGuidesHow to check if you block GPTBot or ClaudeBot — five minutes
Guides

How to check if you block GPTBot or ClaudeBot — five minutes

Plenty of sites serve browsers fine but return 403 to AI crawlers. Here's how to check GPTBot, ClaudeBot and PerplexityBot access yourself with one curl line.

When we measured ten K-beauty exporters on one day, three were returning 403 to ChatGPT's crawler (GPTBot). They loaded fine in a browser but shut the door on AI — and none of the three knew. The usual cause is a misfiring CDN security rule. The good news: checking takes five minutes and fixing takes a few config lines.

Why it's fatal

If GPTBot is blocked, ChatGPT cannot pull a single fact from that brand's official site. When an overseas shopper asks "what's this brand like?", the AI builds its answer only from third parties — retailer pages, old blogs, forums. The brand loses control of its own story. This is exactly what our "technical access" dimension measures.

How to check it yourself (curl)

Not comfortable with a terminal? Stop here. What follows is the developer/marketer route. If the commands feel out of reach, request a free scorecard and we'll check your site's crawler access for you and report back within two days — you can just receive the result.

If you do use a terminal, it's two lines. First, pretend to be a normal browser:

curl -sI -A "Mozilla/5.0" https://yourbrand.com/ | head -1

Then pretend to be GPTBot and compare:

curl -sI -A "Mozilla/5.0 (compatible; GPTBot/1.0; +https://openai.com/gptbot)" https://yourbrand.com/ | head -1

If the first line is 200 but the second is 403 or 429, you're blocked. Repeat with the other user-agents below — just swap the UA token.

CrawlerUser-agent tokenFor
GPTBotGPTBotChatGPT training/search
OAI-SearchBotOAI-SearchBotChatGPT search citations
ClaudeBotClaudeBotClaude
PerplexityBotPerplexityBotPerplexity

Check robots.txt too

Sometimes the block is an explicit robots.txt rule rather than the CDN. Open https://yourbrand.com/robots.txt; if you see Disallow: / under User-agent: GPTBot, that's the cause — deleting a few lines fixes it.

If you find a block

On a CDN (e.g. Cloudflare), add the four AI crawlers to the allow list in your bot rules. For a robots.txt issue, remove the Disallow line. Both are instant fixes, which is why our reports always rank them first. Re-run the curl above and confirm a 200. This is the same method we used for the K-Beauty Index.

Next step

Want this measured on your own site?

The free scorecard runs these checks against your live pages and tells you which of the three pillars is your bottleneck — in two business days, no charge.

Get a free scorecard →