In short. Anthropic asks URL blocking requests to include evidence that the sender owns the URL.
Definition: Anthropic describes a Claude search URL blocking request as an email for a URL that already appears in Claude outputs. The request should include information that proves you own the URL.
Key facts
- Anthropic's content removal guidance separates a URL request from controls such as removing site content, password protection, a
noindextag, and robots.txt rules. - The URL request goes to
webresultsoptout@anthropic.com. - Anthropic says the email should include information needed to prove ownership of the URL.
- The listed examples are an ICANN registration, a WHOIS lookup result, a domain registration payment receipt, an SSL certificate, or a domain-specific email address.
- Anthropic's privacy documentation identifies Claude-SearchBot as the robot used to improve search result quality for users.
What request does Anthropic document?
The relevant procedure applies when a URL already appears in Claude outputs that use web search. Anthropic places it in a table under “Block or remove websites from Claude web search”, with “Applicable: all content types”. The table labels this route “Block access to a URL already appearing in Claude outputs”.
Send the request to the mailbox shown in the guidance:
webresultsoptout@anthropic.com
The email needs information that proves ownership of the URL. The guidance presents the evidence as examples rather than as a form with mandatory fields. It does not state whether one listed item is sufficient or give a processing time.
Which ownership evidence can you provide?
Anthropic lists five forms of evidence: an ICANN registration, a WHOIS lookup result, a domain registration payment receipt, an SSL certificate, or a domain-specific email address. The same guidance links “ICANN registration” to ICANN's page about registering domain names and links “WHOIS lookup result” to a WHOIS service.
The evidence should identify the domain or URL named in the request. A domain-specific email address means an address using the domain that the request concerns. An SSL certificate is listed as evidence in its own right. The page does not prescribe a file format, naming convention, or attachment size for these materials.
How does this differ from a robots.txt control?
Anthropic documents robots.txt as a separate site-wide crawler control. Its privacy documentation names three robots: ClaudeBot for collecting web content that could contribute to model training, Claude-User for retrieving content in response to a user's query, and Claude-SearchBot for search result quality. The documentation says Anthropic's bots honor industry-standard directives in robots.txt.
For a whole-website opt-out, the documented example is placed in the top-level robots.txt file and applies to each subdomain that the site owner wants to opt out:
User-agent: ClaudeBot
Disallow: /
That example addresses crawling by a bot across a website. The email route addresses a URL that is already appearing in Claude outputs, so the two instructions concern different controls.
How should a site owner prepare the request?
Identify the exact URL that appears in Claude outputs, then send the request from or with evidence tied to its domain. Attach or reference one or more of Anthropic's listed ownership records, and use the URL blocking mailbox rather than the separate reporting form for content that violates usage policies or local laws. Anthropic's guidance reserves that form for reports about content in Claude outputs or shared Claude content.
The privacy documentation gives a separate contact, claudebot@anthropic.com, for questions about Anthropic bots and asks site owners to write from an email that includes the domain being contacted about because it is otherwise difficult to verify reports.
Related terms
FAQ
Is a robots.txt rule the URL blocking request?
No. Anthropic documents the URL request as an email to webresultsoptout@anthropic.com. The robots.txt instructions describe controls for Anthropic bots across a website or its subdomains.
Which evidence should a business send?
Anthropic lists an ICANN registration, a WHOIS lookup result, a domain registration payment receipt, an SSL certificate, or a domain-specific email address. The request should identify the URL whose ownership is being shown.
Should the request use the content reporting form?
The guidance assigns the standalone form to content that violates Anthropic's usage policies or local laws and to content in Claude outputs or shared Claude content. A request to block a URL already appearing in Claude outputs uses the separate email address.
Sources
- Claude Help Center, "Report, block, and remove content from Claude", 4 September 2026
- Anthropic Privacy Center, "Does Anthropic crawl data from the web, and how can site owners block the crawler?", 7 April 2026
Check how a page reads to answer engines. Open the GEO score analyzer