Recently, a number of studies have shown that chatbots are outperforming human doctors when it comes to empathy. Insofar as empathy is recognized as a core value of healthcare, this seems promising. But chatbots cannot actually be empathetic: they have no agency, no intentionality, nor any concept of empathy. We may ask ourselves if this really matters. Perhaps the beneficial effects of empathy in the healthcare context can be achieved when an interaction is only perceived as empathetic, without there being any intention of empathy. But this would be a technosolutionist fallacy, whereby our technological solution – seemingly empathetic chatbots – leads to a redefinition of our concept of empathy – a sequence of keywords in a text that does not feel rushed. In the process, the real problem at stake in this context, that healthcare professionals do not have enough time to practice empathetic care, will remain unsolved, becoming an “orphan problem”.