Anthropic bans fake source networks that pollute AI answers

In its updated usage policy, Anthropic has banned using Claude to seed the sources that search engines or AI systems draw answers from with content that misrepresents its origin, authorship or independence. According to Search Engine Journal, the policy's example is a network of sites that pose as unrelated sources and back the same claims.

What exactly does the policy prohibit?

The policy also bars fake personas, outlets and reviews, spreading content through ostensibly independent sites or accounts to hide a shared source, and selling these activities as a service. According to Anthropic, these rules were previously scattered across different parts of the policy; the new section covers deceptive activity of any kind.

How does it relate to Google's rules?

Google's spam policies treat attempts to manipulate generative AI responses in Google Search as spam. The difference is in enforcement: Google applies its rules to websites, while Anthropic applies its policy to Claude users. Anthropic says it may suspend or terminate access if it suspects a violation.

What should brands and agencies do?

For teams working on visibility in AI answers (GEO), the message is clear: building a source network may look attractive in the short term, but search engine and AI provider rules now target it explicitly. As an interpretation, the sustainable route is to become a source on the brand's own site, with real authors and verifiable information.

• Ask your agency or content supplier in writing who owns the sites it publishes on and who the authors are.
• Remove from your inventory, or transparently link to your brand, any network of sites and accounts that presents the same claim as if from independent sources.
• Collect reviews and references only from real customers.
• Keep author, source and publisher information open on AI-assisted content.