Anthropic says some Claude models can now end ‘harmful or abusive’ conversations 

By admin

Anthropic says new capabilities allow its latest AI models to protect themselves by ending abusive conversations.