El País blocks GPTBot, ClaudeBot and PerplexityBot
The wire’s crawler access panel caught El País’s homepage flip from a 200 to a 403 for GPTBot, ClaudeBot and PerplexityBot alike between the September 13 and 16 captures, verified still live at curation time. The Spanish daily’s robots.txt already disallowed ClaudeBot and PerplexityBot by name; GPTBot was not named there, so its block is new at the enforcement layer even though the wildcard rule never covered it explicitly. Separately, Stack Overflow moved Anthropic’s ClaudeBot off Cloudflare’s Pay Per Crawl track — a 402 response offering paid access — onto a flat 403 block, while GPTBot and PerplexityBot stayed behind the same Cloudflare challenge they hit before.
Why it matters: AI-crawler access keeps moving at the enforcement layer, not just in robots.txt — El País shut out all three major crawlers at once, and Stack Overflow ended a paid-access arrangement for one of them, changes neither site's robots.txt alone would show.
The record: CloudflareAnthropicOpenAI
Posted to the wire September 16, 2026. Edited by Joe Balewski.