Weekly report · 2017-W30

The state of AI crawling

Measured on 30,462 reachable domains for the week ending 2017-07-30. 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.

← 2017-W29 · 2017-W31 →

30,462

domains measured

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

Crawler	Operator	Content used for	Block rate	Sites	Week on week
GPTBot	OpenAI	training	11.3%	3,437 of 30,460	new
CCBot	Common Crawl Foundation	archive	11.1%	3,374 of 30,460	new
Bytespider	ByteDance	training	10.7%	3,253 of 30,460	new
ClaudeBot	Anthropic	training	9.9%	3,017 of 30,460	new
Meta-ExternalAgent	Meta	training	8.9%	2,722 of 30,460	new
Google-Extended	Google	training	8.9%	2,718 of 30,460	new
Amazonbot	Amazon	training	8.9%	2,712 of 30,460	new
Applebot-Extended	Apple	training	8.7%	2,650 of 30,460	new
anthropic-ai	Anthropic	training	4.2%	1,269 of 30,460	new
omgilibot	Webz.io	training	4.1%	1,259 of 30,460	new
Diffbot	Diffbot	training	3.8%	1,169 of 30,460	new
cohere-ai	Cohere	live retrieval	3.8%	1,150 of 30,460	new
omgili	Webz.io	training	3.8%	1,144 of 30,460	new
SemrushBot-OCOB	Semrush	training	3.7%	1,133 of 30,460	new
PerplexityBot	Perplexity	answer index	3.6%	1,092 of 30,460	new
ChatGPT-User	OpenAI	live retrieval	3.6%	1,090 of 30,460	new
FacebookBot	Meta	training	3.2%	983 of 30,460	new
YouBot	You.com	answer index	2.9%	880 of 30,460	new
ImagesiftBot	ImageSift (Hive)	training	2.7%	834 of 30,460	new
Timpibot	Timpi	training	2.7%	831 of 30,460	new
Scrapy	Zyte (open-source framework)	training	2.4%	721 of 30,460	new
AI2Bot	Allen Institute for AI	training	2.2%	685 of 30,460	new
OAI-SearchBot	OpenAI	answer index	2.2%	682 of 30,460	new
cohere-training-data-crawler	Cohere	training	2.0%	597 of 30,460	new
DuckAssistBot	DuckDuckGo	live retrieval	1.7%	526 of 30,460	new
Meta-ExternalFetcher	Meta	live retrieval	1.7%	525 of 30,460	new
PanguBot	Huawei	training	1.7%	513 of 30,460	new
Claude-SearchBot	Anthropic	answer index	1.6%	478 of 30,460	new
Claude-User	Anthropic	live retrieval	1.5%	469 of 30,460	new
Kangaroo Bot	Kangaroo LLM	training	1.5%	459 of 30,460	new
Perplexity-User	Perplexity	live retrieval	1.5%	443 of 30,460	new
MistralAI-User	Mistral AI	live retrieval	1.4%	424 of 30,460	new
Applebot	Apple	answer index	1.1%	348 of 30,460	new
Google-CloudVertexBot	Google	live retrieval	1.1%	348 of 30,460	new
Meta-WebIndexer	Meta	answer index	1.1%	346 of 30,460	new
GoogleOther	Google	training	1.1%	330 of 30,460	new
TikTokSpider	ByteDance	live retrieval	0.8%	255 of 30,460	new
Amzn-SearchBot	Amazon	answer index	0.7%	216 of 30,460	new
Amzn-User	Amazon	live retrieval	0.4%	126 of 30,460	new
MistralAI-Index	Mistral AI	answer index	0.2%	69 of 30,460	new
facebookexternalhit	Meta	live retrieval	0.2%	57 of 30,460	new
MistralAI-Training	Mistral AI	training	0.2%	57 of 30,460	new
Bingbot	Microsoft	answer index	0.1%	41 of 30,460	new
Googlebot-News	Google	answer index	0.1%	39 of 30,460	new
msnbot	Microsoft	answer index	0.1%	27 of 30,460	new
Meta-ExternalAds	Meta	training	0.1%	20 of 30,460	new
Googlebot	Google	answer index	0.1%	16 of 30,460	new
OAI-AdsBot	OpenAI	live retrieval	0.0%	3 of 30,460	new

Trend share of measured sites

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

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, 2017-W30", 2017-07-30. https://crawlcensus.com/report/2017-W30

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