Claude Opus 4 and 4.1 Can Now End Abusive Chats on Their Own

Anthropic has given Claude Opus 4 and 4.1 the ability to terminate conversations in its consumer chat interfaces without user input. The feature triggers only in rare, extreme cases of persistently harmful or abusive interactions and is framed as a safety and model-welfare safeguard.

• Claude Opus 4 and 4.1 can end a chat without user input
• Shutdown triggers only on persistent harmful or abusive behavior
• Anthropic calls it a rare, extreme-case safeguard
• Company admits uncertainty about the moral status of LLMs