Computer says "No": The Case Against Empathetic Conversational AI

It is important to note that Recent work in conversational AI has focused our argument applies to any use of empathetic AI on generating empathetic responses to users' (see also for example (Morris et al., 2018; De Carolis emotional states (e.g., Ide and Kawahara, 2022; et al., 2017)). What happens if the chatbot gets Svikhnushina et al., 2022; Zhu et al., 2022) as a way it right? There may be instances where a chatbot to increase or maintain engagement and rapport correctly identifies that a given situation is worthy with the user and to simulate intelligence. However, of praise and amplifies the pride of the user and these empathetic responses are problematic.