Weekly report · 2003-W12

The state of AI crawling

Measured on 30,405 reachable domains for the week ending 2003-03-23. 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.

← 2003-W11  ·  2003-W13 →

30,405
domains measured
7%
block an answer engine
20%
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 7.2% 2,203 of 30,402 new
CCBot Common Crawl Foundation archive 6.9% 2,092 of 30,402 new
Bytespider ByteDance training 6.6% 2,013 of 30,402 new
ClaudeBot Anthropic training 5.9% 1,798 of 30,402 new
Amazonbot Amazon training 5.0% 1,510 of 30,402 new
Applebot-Extended Apple training 4.9% 1,488 of 30,402 new
Meta-ExternalAgent Meta training 4.9% 1,487 of 30,402 new
Google-Extended Google training 4.8% 1,471 of 30,402 new
anthropic-ai Anthropic training 4.5% 1,381 of 30,402 new
omgilibot Webz.io training 4.5% 1,375 of 30,402 new
Diffbot Diffbot training 4.3% 1,310 of 30,402 new
cohere-ai Cohere live retrieval 4.3% 1,296 of 30,402 new
omgili Webz.io training 4.2% 1,262 of 30,402 new
SemrushBot-OCOB Semrush training 3.7% 1,113 of 30,402 new
PerplexityBot Perplexity answer index 3.4% 1,037 of 30,402 new
ChatGPT-User OpenAI live retrieval 3.4% 1,027 of 30,402 new
FacebookBot Meta training 3.1% 951 of 30,402 new
YouBot You.com answer index 3.0% 897 of 30,402 new
ImagesiftBot ImageSift (Hive) training 2.8% 839 of 30,402 new
Timpibot Timpi training 2.7% 824 of 30,402 new
Scrapy Zyte (open-source framework) training 2.4% 725 of 30,402 new
AI2Bot Allen Institute for AI training 2.3% 708 of 30,402 new
OAI-SearchBot OpenAI answer index 2.1% 626 of 30,402 new
cohere-training-data-crawler Cohere training 2.0% 619 of 30,402 new
Meta-ExternalFetcher Meta live retrieval 1.7% 527 of 30,402 new
DuckAssistBot DuckDuckGo live retrieval 1.7% 523 of 30,402 new
PanguBot Huawei training 1.7% 517 of 30,402 new
Google-CloudVertexBot Google live retrieval 1.7% 514 of 30,402 new
Claude-User Anthropic live retrieval 1.7% 504 of 30,402 new
Claude-SearchBot Anthropic answer index 1.6% 480 of 30,402 new
Perplexity-User Perplexity live retrieval 1.6% 480 of 30,402 new
GoogleOther Google training 1.6% 474 of 30,402 new
MistralAI-User Mistral AI live retrieval 1.5% 469 of 30,402 new
Kangaroo Bot Kangaroo LLM training 1.5% 454 of 30,402 new
Meta-WebIndexer Meta answer index 1.2% 363 of 30,402 new
Applebot Apple answer index 1.2% 358 of 30,402 new
TikTokSpider ByteDance live retrieval 0.9% 261 of 30,402 new
MistralAI-Training Mistral AI training 0.8% 254 of 30,402 new
Amzn-SearchBot Amazon answer index 0.8% 248 of 30,402 new
Amzn-User Amazon live retrieval 0.6% 185 of 30,402 new
MistralAI-Index Mistral AI answer index 0.3% 94 of 30,402 new
facebookexternalhit Meta live retrieval 0.2% 48 of 30,402 new
Googlebot-News Google answer index 0.1% 37 of 30,402 new
Bingbot Microsoft answer index 0.1% 35 of 30,402 new
Meta-ExternalAds Meta training 0.1% 24 of 30,402 new
Googlebot Google answer index 0.1% 18 of 30,402 new
msnbot Microsoft answer index 0.1% 18 of 30,402 new
OAI-AdsBot OpenAI live retrieval 0.0% 12 of 30,402 new

Trend share of measured sites

% of sites20151050% of sites08-2309-2110-21block 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=449)16.5%.de (n=842)16.3%.fr (n=333)15.6%.jp (n=423)9%.com (n=14889)6.9%.br (n=374)6.4%.org (n=1665)5.7%.net (n=1310)4.7%.in (n=310)2.9%.ru (n=1154)2.7%.edu (n=465)2.4%.io (n=658)2.1%

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, 2003-W12", 2003-03-23. https://crawlcensus.com/report/2003-W12

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