Anthropic Bans Cruel Behavior Toward Claude AI Models

Anthropic updated its usage policy on October 8, 2026, to prohibit users from exhibiting sustained and needless abusive or cruel behavior toward its AI models, including the chatbot Claude. Effective November 12, the company will treat extreme cases of repeated cruelty without a discernible purpose as policy violations, though it clarified the rule does not apply to common user frustration or research testing. Enforcement will primarily rely on Claude's ability to end conversations, with further penalties including warnings, throttling, or account termination.

The policy update coincides with broader restrictions on deceptive activities, such as the use of fake accounts for influence operations, and a ban on developing weapons software or non-consensual surveillance tools. These surveillance restrictions are central to an ongoing legal dispute between Anthropic and the United States Department of Defense regarding the government's use of models for mass surveillance.

The move has triggered a philosophical debate over AI sentience. While Anthropic does not explicitly cite model welfare, CEO Dario Amodei stated he is open to the idea that AI models could be conscious. Conversely, Microsoft AI chief Mustafa Suleyman and OpenAI CEO Sam Altman rejected the notion of AI consciousness, with Suleyman warning that granting moral protections to technological entities is a recipe for disaster. Pope Leo XIV also asserted in a sermon that machines lack a soul and merely compile data.