A study of how 3 popular artificial intelligence chatbots respond to queries about suicide found that they generally avoid answering questions that pose the highest risk to the user, such as for specific how-to guidance.
News for you 247