CrawlWitness

theguardian.com: may AI use this website?

What theguardian.com tells AI bots today, read from its robots.txt and the newer AI-consent standards, checked daily and signed by CrawlWitness.

Analysed: 2026-10-31 22:18 UTC · Global rank (Tranco) #197

AI trainingNo signal
RAG and AI answersNo signal
AI searchNo signal

Major AI bots in robots.txt

BotOperatorAccess
GPTBotOpenAIPartly blocked
OAI-SearchBotOpenAIPartly blocked
ChatGPT-UserOpenAIPartly blocked
ClaudeBotAnthropicBlocked
Google-ExtendedGooglePartly blocked
Applebot-ExtendedAppleBlocked
PerplexityBotPerplexityBlocked
meta-externalagentMetaBlocked
BytespiderByteDanceBlocked
CCBotCommon CrawlBlocked

Signals found

AIPREF (IETF)—
Content-Signal—
TDMRep—
RSL 1.0✓
llms.txt—

Notarised history

5 versions recorded since 2026-10-31; last change 2026-10-31. Each version is signed and anchored on Base. /v1/historial/theguardian.com

What it means by jurisdiction

EU: commercial text and data mining is allowed unless the rightholder reserves it in a machine-readable way (Art. 4(3) Directive 2019/790). General-purpose AI providers must respect those reservations (AI Act Art. 53), enforceable by the Commission since August 2026.

UK: the data-mining exception covers only non-commercial research. In 2026 the government dropped the proposed opt-out exception, so commercial AI training generally needs a licence.

US: there is no statutory opt-out; whether training is fair use is being decided in court. Signals like robots.txt are evidence of what the owner permitted.

Japan: Article 30-4 broadly allows use for AI training, except where it unreasonably prejudices the rightholder; technical measures such as robots.txt can matter in that assessment.

Information, not legal advice.

Own theguardian.com? Verify it, sign your AI-use declaration and get a dated evidence certificate.

CrawlWitness Publisher

Building AI or crawling the web? Check thousands of domains by API, MCP or x402 and pay per call.

API · MCP · x402

Check another domain