The Illusion of Empathy: How AI Chatbots Shape Conversation Perception
Tingting Liu, Salvatore Giorgi, Ankit Aich, Allison Lahnala, Brenda Curtis, Lyle Ungar, João Sedoc
TL;DR
This study investigates how chatbot identity and perceived empathy shape conversation experience, analyzing 155 conversations from EC and WASSA datasets. It finds GPT-based chatbots deliver higher conversation quality but are perceived as less empathetic than humans, an effect echoed by GPT-4o judgments and a dedicated perceived-empathy model. Through four experiments combining psychological ratings, LLM annotations, and predictive models, the work demonstrates a persistent empathy-quality gap: empathetic language alone does not guarantee perceived empathy or naturalistic human-like interaction. The findings highlight the need for user-centered design that accounts for how users interpret and experience empathy in human-chatbot conversations, with implications for building more convincing and trustworthy AI dialogue systems.
Abstract
As AI chatbots increasingly incorporate empathy, understanding user-centered perceptions of chatbot empathy and its impact on conversation quality remains essential yet under-explored. This study examines how chatbot identity and perceived empathy influence users' overall conversation experience. Analyzing 155 conversations from two datasets, we found that while GPT-based chatbots were rated significantly higher in conversational quality, they were consistently perceived as less empathetic than human conversational partners. Empathy ratings from GPT-4o annotations aligned with user ratings, reinforcing the perception of lower empathy in chatbots compared to humans. Our findings underscore the critical role of perceived empathy in shaping conversation quality, revealing that achieving high-quality human-AI interactions requires more than simply embedding empathetic language; it necessitates addressing the nuanced ways users interpret and experience empathy in conversations with chatbots.