The AI bots, ranked by unwelcome

Every AI-associated user agent we check, and how many of the 646 readable sites block it. Snapshot: 2026-11-15

BotOperatorPurposeBlocked by
GPTBotOpenAItraining142 (22%)
ChatGPT-UserOpenAIfetch92 (14.2%)
OAI-SearchBotOpenAIsearch77 (11.9%)
ClaudeBotAnthropictraining149 (23.1%)
Claude-UserAnthropicfetch102 (15.8%)
Claude-SearchBotAnthropicsearch99 (15.3%)
Google-ExtendedGoogletraining122 (18.9%)
Applebot-ExtendedAppletraining131 (20.3%)
PerplexityBotPerplexitysearch120 (18.6%)
Perplexity-UserPerplexityfetch107 (16.6%)
CCBotCommon Crawltraining163 (25.2%)
BytespiderByteDancetraining166 (25.7%)
meta-externalagentMetatraining138 (21.4%)
AmazonbotAmazontraining122 (18.9%)
DuckAssistBotDuckDuckGosearch98 (15.2%)
MistralAI-UserMistralfetch97 (15%)
anthropic-aiAnthropictraining121 (18.7%)
Claude-WebAnthropicfetch119 (18.4%)
cohere-aiCoheretraining133 (20.6%)
cohere-training-data-crawlerCoheretraining102 (15.8%)
DiffbotDiffbottraining141 (21.8%)
omgiliWebz.iotraining135 (20.9%)
omgilibotWebz.iotraining117 (18.1%)
YouBotYou.comsearch114 (17.6%)
ImagesiftBotHivetraining116 (18%)
TimpibotTimpisearch114 (17.6%)
Meta-ExternalFetcherMetafetch98 (15.2%)
AI2BotAllen Institute for AItraining101 (15.6%)
DeepSeekBotDeepSeektraining81 (12.5%)
GrokBotxAItraining70 (10.8%)
PanguBotHuaweitraining96 (14.9%)
Gemini-Deep-ResearchGooglefetch72 (11.1%)

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.