The question
A systematic review of how empathy has been conceptualized, designed, and evaluated in research on conversational agents.
What we learned
The review treats empathy as context-dependent: design goals, conversational strategies, and evaluation methods vary across health, service, social, and learning settings. It identifies gaps in longitudinal evaluation and voice-based interfaces. The paper received an Honorable Mention at ACM CUI 2026.
Part of CHAI Lab’s work on support through conversational AI.