Cultivating Compassion: How GenAI Chatbots Enhance Human Counselors' Empathy

Abstract

Telehealth platforms increasingly integrate GenAI chatbots into free services to expand capacity, yet how this practice influences human care providers in adjacent paid modules remains unclear. Using data from two leading Chinese teletherapy platforms, we find that introducing GenAI chatbots into the free module significantly increases human counselors' paid orders and improves care seeker satisfaction in the paid module. To understand the mechanism driving these effects, we examine how counselors respond to chatbot deployment and test two potential pathways: direct competition and differentiation. Direct competition would involve counselors increasing activity in the free module to compete for care seeker attention. We find no evidence of such behavior. Instead, counselors pursue differentiation-they enhance multiple dimensions of empathic expression during paid consultations, leveraging their comparative advantage in providing authentic emotional connection. This empathy enhancement subsequently drives performance gains. Heterogeneity analyses reveal stronger effects for counselors with lower baseline empathy levels and those facing higher competitive pressure, as well as for counseling topics characterized by high emotional intensity, complex psychological states, and greater demand for empathy. By documenting cross-module spillover effects of GenAI chatbots on teletherapy platforms, our study demonstrates how GenAI can facilitate the efficient delivery of compassionate care. For practitioners, our findings suggest that strategic deployment of GenAI chatbots can strengthen value creation on telehealth platforms by augmenting, rather than replacing, human services.

Keywords: Compassionate Care, Mental Health, Human-AI Collaboration, Freemium Model, Cross Module