Check
ChatGPT cannot read my site
An AI assistant answers about your site from memory, or says it cannot access it. The page returns 200 in your browser, so nothing looks wrong — the request is being refused for the crawler's user agent.
How to confirm it
Run this against your own domain:
curl -sS -A "GPTBot/1.0 (+https://openai.com/gptbot)" -o /dev/null -w "%{http_code}\n" https://example.com/
How to fix it
A 403 or 406 means a WAF, CDN or bot rule matched the agent. Allowlist GPTBot, ClaudeBot, PerplexityBot and OAI-SearchBot by user agent, and remove the JavaScript challenge for them: a challenge page returns 200 with no content, which reads as an empty site.
Where it fits
This is a symptom of a failed check — here it points at A1. The failure it belongs to is only one of fifteen; the full list is in the standard, and the fix order that avoids redoing work is in the repair guide.