Empathetic language in AI travel services: the moderating effects of system adaptivity and AI disclosure on user intentions

This research investigates how empathetic language influences user engagement in AI-driven travel services, focusing on the roles of AI adaptivity and disclosure. Grounded in the Elaboration Likelihood Model, two experimental studies demonstrate that empathetic language enhances user intentions when AI adaptivity is low; and that cognitive empathy is more persuasive than affective empathy when AI is disclosed. These findings challenge the belief that humanlike AI is always better and emphasise the need to match empathy with system traits and user expectations. This study contributes to the hospitality AI literature by showing when and how empathy is effective in AI communication.