anthropic-ai: robots token, verification and how to allow or disallow it

All 21 documented crawlers

Identity

documentation_url is null on purpose: the vendor help article fetched on 2026-09-06 documents three robots (ClaudeBot, Claude-User, Claude-SearchBot) and contains zero occurrences of 'anthropic-ai'. The token is still widely present in site robots.txt files, so it stays in the registry as a legacy token whose current behaviour the vendor does not document.

Allow or disallow it

User-agent: anthropic-ai
Allow: /

User-agent: anthropic-ai
Disallow: /

A robots.txt rule is a request that a well behaved crawler honours. It is not an access control, and this page does not tell anyone what to choose.

Build a robots.txt with anthropic-ai and every other documented token

Questions

How do I allow anthropic-ai in robots.txt?

Add this group to the robots.txt at the root of the host: User-agent: anthropic-ai Allow: /. The token is matched case insensitively as a substring of the user agent by RFC 9309, and the longest matching rule wins.

How do I disallow anthropic-ai in robots.txt?

Add this group to the robots.txt at the root of the host: User-agent: anthropic-ai Disallow: /. A robots.txt rule is a request that a well behaved crawler honours; it is not an access control.

How do I verify a request really came from anthropic-ai?

We have not found a documented verification method for anthropic-ai: the vendor publishes neither an IP range file nor a reverse DNS convention that we could fetch. Without one, a user agent string is not evidence of origin.

What we measure

Of the 12976 seeded domains whose robots.txt we have read, 1277 disallow anthropic-ai for / and 11699 allow it, which is 9.8% disallowed, week 2026-W43. Counts only: no domain is named.

The weekly AI crawler access index

Sources

Free data

• The AEO Watch dataset
• Tranco top 1,000 AI crawler policy (CSV, CC BY 4.0)
• The AEO Watch API
• The checker as a command line tool (one file, no install)

Check a domain against every token

Check your own site

The same checks, run on your domain now: what robots.txt tells each AI crawler, whether the edge answers them, and whether the text is readable without JavaScript. Free, no signup.

Limits: 8 checks and 5 different domains per address per hour; a repeat of the same domain inside 7 days is answered from what we already hold.

AEO Watch is an independent, factual monitor. It is not affiliated with, endorsed by or speaking for any crawler vendor. Every statement is an observation with the date it was made and the raw evidence behind it: a robots.txt line, an HTTP status code, a header. There are no scores, no grades and no verdicts here, and nothing on this page is advice. A site opts out at any time and the opt out is honoured automatically and permanently.