Anthropic's usage policy bars seeding fake sources to sway AI answers

Anthropic’s usage policy, effective November 12, 2026, now prohibits using Claude to “manipulate the sources from which search engines or AI systems draw answers by seeding them with content that misrepresents its origin, authorship, or independence” — language absent from the prior version. The same update drops “automatically generate content and publish it for external consumption” from the policy’s high-risk use cases, which had required professional review before publication; Anthropic has not explained the removal, per a comparison of the current and archived policy text. The new ban follows Anthropic’s September 2026 threat intelligence report, which identified roughly 70 fake local news outlets, run by a France-based ad agency, that used Claude to publish more than 8,900 articles across 20 languages designed to manipulate search rankings.