Anthropic prohibits using Claude to create misleading search sources
October 12, 2026 · 2 min read
Anthropic, the developer of the AI assistant Claude, has published updated usage rules for its service. These rules explicitly prohibit the creation of websites that present themselves as independent sources to influence the responses of search engines and AI. The new policy will take effect on November 12, 2026.
Indicators of misleading campaigns include fake accounts, organizations, media, and reviews, undisclosed payment for publications, and the spread of a single message through interconnected platforms. A separate clause addresses the content of sources relied upon by search engines and generative models, including materials with false authorship or claims of independence.
For specialists in SEO, SERM, and GERM, the boundary of the prohibition's application is crucial. Anthropic regulates the use of its own products, including Claude and its API. This does not imply that search engines have changed their ranking algorithms or that there are mechanisms that automatically detect such networks of sites.
Anthropic explains that these restrictions existed previously but were scattered across sections concerning fraud, disinformation, and elections. The company has now consolidated them into a separate category that encompasses not only political but also commercial campaigns. Therefore, it is not about banning ordinary content automation, but rather cases of intentional distortion of the source of origin of content and the scale of its dissemination.
The main takeaway for promotion through generative search is narrow: using Claude to simulate pseudo-independent confirmations is now explicitly named a violation of the platform's rules. The announcement does not disclose how often and in what manner the provider will identify such cases.