Anthropic has added a clause to its usage policy prohibiting “sustained and needless abusive or cruel behavior toward our models.” The rule, announced on October 8, 2026, and effective November 12, 2026, targets only extreme cases of repeated cruelty with no discernible purpose. Legitimate testing and research are exempt, as are everyday expressions of frustration and creative scenarios. Claude will be able to end conversations it judges to be abusive, a capability introduced in August 2025, and Anthropic may also take broader account-level action. According to Searchable, 33 percent of US chatbot users have sworn at or insulted their AI models over the past year, with 19.1 derogatory words for every 10,000 words sent.
Source: Read the original article
