Machines that express emotions: from medical paternalism to machine paternalism?

Abstract: Introduction: Machines do not have emotions, but they can be programmed to express (to appear to have) certain emotions. In medicine, robots and AI machines are used that could express positive emotions (joy, compassion, hope, gratitude, inspiration or awe) and priorities. Objective: To describe the main ethical implications of integrating the expression of emotions in machines used for health care, especially when they use AI. Results: Some ethical problems generated by machines applied to medicine that integrate the expression of emotions and priorities are: 1) The equivocal “he understands me”, thinking that the machine empathizes, with the corresponding risk of confusion and manipulation; 2) The robot friend, the creation of humanoids that interact with humans, being able to transfer specifically human responsibilities to machines; 3) Overconfidence in machines, putting at risk the privacy of people; 4) Confusing who is the moral subject of medical decisions, which can never be a machine. Conclusions: The answer to these moral challenges involves making appropriate technological designs, because companies and programmers are responsible for the development of their machines, and for making responsible use of the machines by professionals and users. We must not move from the old medical paternalism to a future machine paternalism, bypassing patient autonomy.