Weekly report · 2013-W38

The state of AI crawling

Measured on 30,286 reachable domains for the week ending 2013-09-22. 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.

← 2013-W37  ·  2013-W39 →

30,286
domains measured
6%
block an answer engine
19%
refuse AI agents at the edge
13%
publish an llms.txt

Block rate by crawler with change on the previous week

CrawlerOperatorContent used forBlock rateSitesWeek on week
GPTBot OpenAI training 7.8% 2,367 of 30,288 new
CCBot Common Crawl Foundation archive 7.6% 2,314 of 30,288 new
Bytespider ByteDance training 7.2% 2,180 of 30,288 new
ClaudeBot Anthropic training 6.4% 1,938 of 30,288 new
Amazonbot Amazon training 5.5% 1,662 of 30,288 new
Meta-ExternalAgent Meta training 5.4% 1,650 of 30,288 new
Google-Extended Google training 5.3% 1,614 of 30,288 new
Applebot-Extended Apple training 5.2% 1,588 of 30,288 new
anthropic-ai Anthropic training 4.2% 1,261 of 30,288 new
omgilibot Webz.io training 4.1% 1,241 of 30,288 new
cohere-ai Cohere live retrieval 3.9% 1,171 of 30,288 new
Diffbot Diffbot training 3.8% 1,155 of 30,288 new
SemrushBot-OCOB Semrush training 3.8% 1,142 of 30,288 new
omgili Webz.io training 3.7% 1,125 of 30,288 new
PerplexityBot Perplexity answer index 3.5% 1,054 of 30,288 new
ChatGPT-User OpenAI live retrieval 3.4% 1,042 of 30,288 new
FacebookBot Meta training 3.1% 952 of 30,288 new
YouBot You.com answer index 2.9% 876 of 30,288 new
ImagesiftBot ImageSift (Hive) training 2.7% 826 of 30,288 new
Timpibot Timpi training 2.7% 806 of 30,288 new
Scrapy Zyte (open-source framework) training 2.4% 715 of 30,288 new
AI2Bot Allen Institute for AI training 2.2% 662 of 30,288 new
OAI-SearchBot OpenAI answer index 2.1% 641 of 30,288 new
cohere-training-data-crawler Cohere training 2.0% 599 of 30,288 new
Meta-ExternalFetcher Meta live retrieval 1.8% 535 of 30,288 new
DuckAssistBot DuckDuckGo live retrieval 1.7% 508 of 30,288 new
PanguBot Huawei training 1.6% 498 of 30,288 new
Claude-SearchBot Anthropic answer index 1.6% 472 of 30,288 new
Claude-User Anthropic live retrieval 1.6% 472 of 30,288 new
Kangaroo Bot Kangaroo LLM training 1.5% 454 of 30,288 new
Perplexity-User Perplexity live retrieval 1.5% 446 of 30,288 new
MistralAI-User Mistral AI live retrieval 1.4% 428 of 30,288 new
Google-CloudVertexBot Google live retrieval 1.2% 366 of 30,288 new
GoogleOther Google training 1.2% 351 of 30,288 new
Meta-WebIndexer Meta answer index 1.2% 351 of 30,288 new
Applebot Apple answer index 1.1% 336 of 30,288 new
TikTokSpider ByteDance live retrieval 0.8% 256 of 30,288 new
Amzn-SearchBot Amazon answer index 0.7% 217 of 30,288 new
Amzn-User Amazon live retrieval 0.5% 145 of 30,288 new
MistralAI-Training Mistral AI training 0.3% 81 of 30,288 new
MistralAI-Index Mistral AI answer index 0.2% 67 of 30,288 new
facebookexternalhit Meta live retrieval 0.2% 53 of 30,288 new
Bingbot Microsoft answer index 0.1% 39 of 30,288 new
Googlebot-News Google answer index 0.1% 38 of 30,288 new
msnbot Microsoft answer index 0.1% 26 of 30,288 new
Meta-ExternalAds Meta training 0.1% 20 of 30,288 new
Googlebot Google answer index 0.1% 17 of 30,288 new
OAI-AdsBot OpenAI live retrieval 0.0% 6 of 30,288 new

Trend share of measured sites

% of sites20151050% of sites08-2209-0609-22block 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)17.4%.fr (n=327)16.8%.de (n=832)16.2%.jp (n=409)9%.com (n=14828)6.5%.org (n=1668)5.6%.br (n=374)5.3%.net (n=1311)4%.ru (n=1180)3.1%.in (n=312)2.9%.edu (n=475)2.5%.io (n=657)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, 2013-W38", 2013-09-22. https://crawlcensus.com/report/2013-W38

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