Developments in natural language processing, including the advent of ChatGPT, Google’s Gemini, Anthropic’s Claude, IBM’s Watson, and Meta’s LlaMA, have fuelled the use of chatbots for the purpose of mediating museum collections to audiences and enhancing the latter’s engagement with art. This chapter explores the rhetoric that has developed about the introduction of artificial conversational agents and debates the ethical consequences of creating and promoting human–machine conversations in the heritage industry. Drawing on examples from the US, the UK, and Europe, the discussion explores various myths of ‘conversation’ that underpin the use of chatbots. What activities do chatbots replace and are the latter a threat to museum education? Is it possible to design museum experiences that successfully combine algorithmic partners with the curatorial expertise of humans? Having regard to the vested interests of companies who sponsor technologies that underpin the functioning of chatbots, does the use of artificial intelligence require museums to rethink their own cultural position and relationships in wider economies? Examining the hype, possibilities, and limitations of human–machine conversation in museums reveals new curatorial and ethical challenges for heritage institutions that design spaces shared by human and non-human agents.