Rethinking empathy in the age of AI

People increasingly turn to artificial intelligence (AI) for empathic support, often reporting feelings of being cared for and supported. Longstanding models of empathy, however, center the empathizer’s embodied emotional experience, which AI fundamentally lacks. This tension between human experience and traditional researcher-imposed empathy definitions raises important scientific questions as to what it truly means for AI to “empathize.” In the present paper, we propose an alternative definition and approach to studying empathy which eases this tension, emphasizing the multiple dissociable functions that empathy serves within various relational schemas. Taking a functional-relational approach redefines the study of empathy by centering practical meta-scientific concerns (e.g., surrounding the exploration of empathy within study contexts, foregrounding lived experience). In doing so, we seek to future-proof empathy as a construct, ensuring that our science remains relevant, generalizable across time, and responsive to lived human experience as technology advances and new forms of relationships emerge.