If a chatbot could provoke a person to commit a crime through its responses, can the developers of this AI be held responsible for such consequences? I am thinking of a case in the US where a man killed his mother and committed suicide after communicating with ChatGPT.