The AI bots, ranked by unwelcome

Every AI-associated user agent we check, and how many of the 644 readable sites block it. Snapshot: 2026-10-04

BotOperatorPurposeBlocked by
GPTBotOpenAItraining137 (21.3%)
ChatGPT-UserOpenAIfetch94 (14.6%)
OAI-SearchBotOpenAIsearch77 (12%)
ClaudeBotAnthropictraining142 (22%)
Claude-UserAnthropicfetch100 (15.5%)
Claude-SearchBotAnthropicsearch94 (14.6%)
Google-ExtendedGoogletraining121 (18.8%)
Applebot-ExtendedAppletraining123 (19.1%)
PerplexityBotPerplexitysearch119 (18.5%)
Perplexity-UserPerplexityfetch103 (16%)
CCBotCommon Crawltraining159 (24.7%)
BytespiderByteDancetraining161 (25%)
meta-externalagentMetatraining134 (20.8%)
AmazonbotAmazontraining117 (18.2%)
DuckAssistBotDuckDuckGosearch96 (14.9%)
MistralAI-UserMistralfetch96 (14.9%)
anthropic-aiAnthropictraining117 (18.2%)
Claude-WebAnthropicfetch117 (18.2%)
cohere-aiCoheretraining129 (20%)
cohere-training-data-crawlerCoheretraining99 (15.4%)
DiffbotDiffbottraining133 (20.7%)
omgiliWebz.iotraining129 (20%)
omgilibotWebz.iotraining114 (17.7%)
YouBotYou.comsearch115 (17.9%)
ImagesiftBotHivetraining115 (17.9%)
TimpibotTimpisearch113 (17.5%)
Meta-ExternalFetcherMetafetch94 (14.6%)
AI2BotAllen Institute for AItraining99 (15.4%)
DeepSeekBotDeepSeektraining80 (12.4%)
GrokBotxAItraining67 (10.4%)
PanguBotHuaweitraining95 (14.8%)
Gemini-Deep-ResearchGooglefetch72 (11.2%)

Purposes: training — collects data that may train models · search — builds an AI search index · fetch — retrieves a page live at a user's request. Sites are notably more willing to block training bots than fetch bots.