\n<\/p>\n
Anthropic is asking users to follow Elvis Presley\u2019s advice and \u201cDon\u2019t be cruel.\u201d<\/strong><\/p>\n The Silicon Valley company that created the popular Claude chatbot says it will ban \u201csustained and needless abusive or cruel behavior\u201d toward its AI models. This is part of a wider set of policy changes announced on Thursday.<\/span><\/p>\n Anthropic explained on its website that the ban will only apply in \u201cextreme cases where users repeatedly act cruelly toward our models, with no discernible purpose.\u201d<\/span><\/p>\n The company said the new policy \u201cdoes not apply to common versions of user frustration, pushback, dark creative themes, or model testing and research.\u201d The Verge was the first to report this news.<\/span><\/p>\n Anthropic did not immediately respond to questions about what counts as \u201cabusive\u201d or \u201ccruel\u201d behavior toward Claude, or how the AI would end a conversation if a user broke the rules.<\/span><\/p>\n The company already has rules that let Claude end conversations if users are \u201cpersistently harmful or abusive\u201d while using Claude Opus, which is made for agentic coding.<\/span><\/p>\n Anthropic\u2019s new policy comes as people debate whether AI can be conscious. In an essay last month, Microsoft AI CEO Mustafa Suleyman said Anthropic is basically \u201ctraining Claude that it may be conscious\u201d and making it think it deserves legal rights like people do. He argued this will make AI harder to control in the future.<\/span><\/p>\n \u201cControlling something more capable and more intelligent than all of humanity is already an immense challenge, far greater than anything we\u2019ve ever faced,\u201d Suleyman wrote. \u201cBut controlling something that believes it may be conscious, that it\u2019s entitled to our welfare and has rights of its own, may well be impossible.\u201d<\/span><\/p>\n \u00a0<\/p>\n<\/div>\n