Anthropic Bans ‘Cruel’ Behaviour Towards AI Models, Tightens Rules On AI Misuse

Anthropic has revised its guidelines to prohibit abusive behavior towards its AI models, specifically directed at Claude. The restrictions are only enforced in extreme cases where users persistently engage in cruel actions with no apparent motive. The policy update, effective November 12, does not cover typical user frustration or experimentation.

Users can terminate conversations at any time, but Claude will only do so as a last resort after unsuccessful attempts to redirect the interaction. The policy also expands Claude's restrictions to combat AI misuse, including deceptive political campaigns and weapons-related activities.