Weekly report · 2010-W20

The state of AI crawling

Measured on 30,093 reachable domains for the week ending 2010-05-23. Every figure is the result of an automated audit run the same way on every domain: read robots.txt, resolve the rules for 48 tracked agents, then send live requests as four of them.

← 2010-W19  ·  2010-W21 →

30,093
domains measured
7%
block an answer engine
19%
refuse AI agents at the edge
14%
publish an llms.txt

Block rate by crawler with change on the previous week

CrawlerOperatorContent used forBlock rateSitesWeek on week
GPTBot OpenAI training 7.0% 2,115 of 30,092 new
CCBot Common Crawl Foundation archive 6.8% 2,048 of 30,092 new
Bytespider ByteDance training 6.4% 1,928 of 30,092 new
ClaudeBot Anthropic training 5.7% 1,728 of 30,092 new
Meta-ExternalAgent Meta training 4.7% 1,404 of 30,092 new
Amazonbot Amazon training 4.6% 1,390 of 30,092 new
Google-Extended Google training 4.6% 1,371 of 30,092 new
Applebot-Extended Apple training 4.5% 1,355 of 30,092 new
anthropic-ai Anthropic training 4.3% 1,307 of 30,092 new
omgilibot Webz.io training 4.2% 1,272 of 30,092 new
cohere-ai Cohere live retrieval 4.0% 1,207 of 30,092 new
Diffbot Diffbot training 4.0% 1,200 of 30,092 new
omgili Webz.io training 3.8% 1,158 of 30,092 new
SemrushBot-OCOB Semrush training 3.7% 1,114 of 30,092 new
PerplexityBot Perplexity answer index 3.5% 1,039 of 30,092 new
ChatGPT-User OpenAI live retrieval 3.4% 1,038 of 30,092 new
FacebookBot Meta training 3.2% 963 of 30,092 new
YouBot You.com answer index 2.9% 879 of 30,092 new
ImagesiftBot ImageSift (Hive) training 2.8% 843 of 30,092 new
Timpibot Timpi training 2.7% 818 of 30,092 new
Scrapy Zyte (open-source framework) training 2.5% 738 of 30,092 new
AI2Bot Allen Institute for AI training 2.3% 689 of 30,092 new
OAI-SearchBot OpenAI answer index 2.1% 635 of 30,092 new
cohere-training-data-crawler Cohere training 2.0% 606 of 30,092 new
Meta-ExternalFetcher Meta live retrieval 1.8% 547 of 30,092 new
DuckAssistBot DuckDuckGo live retrieval 1.7% 508 of 30,092 new
PanguBot Huawei training 1.7% 507 of 30,092 new
Claude-User Anthropic live retrieval 1.6% 485 of 30,092 new
Claude-SearchBot Anthropic answer index 1.6% 467 of 30,092 new
Kangaroo Bot Kangaroo LLM training 1.5% 458 of 30,092 new
Perplexity-User Perplexity live retrieval 1.5% 456 of 30,092 new
MistralAI-User Mistral AI live retrieval 1.4% 434 of 30,092 new
Google-CloudVertexBot Google live retrieval 1.4% 410 of 30,092 new
GoogleOther Google training 1.2% 372 of 30,092 new
Meta-WebIndexer Meta answer index 1.2% 355 of 30,092 new
Applebot Apple answer index 1.1% 334 of 30,092 new
TikTokSpider ByteDance live retrieval 0.8% 250 of 30,092 new
Amzn-SearchBot Amazon answer index 0.7% 225 of 30,092 new
Amzn-User Amazon live retrieval 0.5% 159 of 30,092 new
MistralAI-Training Mistral AI training 0.4% 130 of 30,092 new
MistralAI-Index Mistral AI answer index 0.2% 74 of 30,092 new
facebookexternalhit Meta live retrieval 0.2% 51 of 30,092 new
Bingbot Microsoft answer index 0.1% 38 of 30,092 new
Googlebot-News Google answer index 0.1% 37 of 30,092 new
Meta-ExternalAds Meta training 0.1% 22 of 30,092 new
msnbot Microsoft answer index 0.1% 21 of 30,092 new
Googlebot Google answer index 0.1% 18 of 30,092 new
OAI-AdsBot OpenAI live retrieval 0.0% 9 of 30,092 new

Trend share of measured sites

% of sites20151050% of sites08-2209-1210-03block an answer engineblock a training crawlerpublish llms.txt

Where the blocking is share blocking an answer engine, by top-level domain

bar chart.uk (n=445)16.4%.de (n=825)15.6%.fr (n=324)14.5%.jp (n=407)8.8%.com (n=14743)6.8%.br (n=364)6.6%.org (n=1666)5.5%.net (n=1301)4.2%.ru (n=1166)3.2%.in (n=313)2.9%.edu (n=473)2.1%.io (n=660)1.8%

Only TLDs with at least 40 measured domains are shown, and each bar carries its sample size. A TLD is a weak proxy for jurisdiction. Full segmentation.

Changes recorded this week 0 total

No crawler policy changes were recorded in this week.

Full change log · RSS

Cite this CC BY 4.0

The figures on this page are frozen to the week shown and will not be edited after publication. Corrections are appended, not overwritten.

Crawl Census, "The state of AI crawling, 2010-W20", 2010-05-23. https://crawlcensus.com/report/2010-W20

Machine-readable: census totals, full corpus, crawler registry. Method: methodology.