Anthropic says new capabilities allow its latest AI models to protect themselves by ending abusive conversations.
About The Author
Anthropic says some Claude models can now end ‘harmful or abusive’ conversations
