Anthropic says some Claude models can now end ‘harmful or abusive’ conversations Leave a Comment / Uncategorized / By Anthropic says new capabilities allow its latest AI models to protect themselves by ending abusive conversations. Related Posts