Anthropic is implementing a policy to prohibit severe abusive behavior toward Claude, its AI chatbot. The rule targets only extreme cases where users repeatedly engage in abusive actions without clear purpose or practical intent. When such behavior occurs, Claude may terminate the conversation, though it remains unclear whether users will face account bans.
The policy explicitly excludes common frustration, disagreements with the model, or exploration of dark themes in creative work. Anthropic is drawing a careful line between ordinary interactions and genuine harassment, focusing enforcement on cases that cross into sustained, purposeless cruelty rather than everyday friction between users and the AI.