Last month, Anthropic invited religious leaders and theologians to its San Francisco headquarters to discuss what values should shape Claude, its chatbot. Company staff asked for advice on how to guide the chatbot’s moral and spiritual development for answering ethical questions from users. The meeting showed how far AI has moved beyond coding and computing. Tech companies are now wrestling with issues philosophers and religions have debated for centuries. Pope Leo XIV has urged developers to cultivate “moral discernment”, while Christian groups are testing chatbots for “theological reliability”. But if religions and cultures all have different ideas of morality, whose beliefs should an AI system follow, and who gets to decide that?
