Claude can now stop conversations – for its own protection, not yours

The company has given its AI chatbot the ability to end toxic conversations as part of its broader ‘model welfare’ initiative.