The company said it would launch tools to safeguard users of its My AI chatbot after reports said it was "was responding in an unsafe and inappropriate manner."
