A new study finds that people value empathy more when they believe it comes from a human—even if the actual response was generated by AI. Across nine studies involving over 6,000 participants, the research reveals that human-attributed responses are perceived as more supportive, more emotionally resonant, and more caring than identical AI-generated responses.
A new international study led by Prof Anat Perry from the Hebrew University of Jerusalem and her PhD student – Matan Rubin, in collaboration with Prof. Amit Goldenberg researchers from Harvard University and Prof. Desmond C. Ong from the University of Texas, finds that people place greater emotional value on empathy they believe comes from humans—even when the exact same response is generated by artificial intelligence.
Published in Nature Human Behaviour, the study involved over 6,000 participants across nine experiments. The researchers tested whether people perceived empathy differently depending on whether it was labeled as coming from a human or from an AI chatbot. In all cases, the responses were crafted by large language models (LLMs), yet participants consistently rated the "human" responses as more empathic, more supportive, and more emotionally satisfying than the identical "AI" responses.
"We're entering an age where AI can produce responses that look and sound empathic," said Prof. Perry. "But this research shows that even if AI can simulate empathy, people still prefer to feel that another human truly understands, feels with them, and cares."
The preference was especially strong for responses that emphasized emotional sharing and genuine care—the affective and motivational components of empathy—rather than just cognitive understanding. In fact, participants were even willing to wait days or weeks to receive a response from a human rather than get an immediate reply from a chatbot.
Interestingly, when participants believed an AI may have helped generate or edit a response they thought was from a human, their positive feelings diminished significantly. This suggests that perceived authenticity—believing that someone genuinely invested time and emotional effort—plays a critical role in how we experience empathy.
"In today's world, it's becoming second nature to run our emails or messages through AI," said Prof. Perry. "But our findings suggest a hidden cost: the more we rely on AI, the more our words risk feeling hollow. As people begin to assume that every message is AI-generated, the perceived sincerity, and with it, the emotional connection, may begin to disappear."
While AI shows promise for use in education, healthcare, and mental health settings, the study highlights its limitations. "AI may help scale support systems," Perry explains, "but in moments that require deep emotional connection, people still want the human touch."
The study offers key insights into the psychology of empathy and raises timely questions about how society will integrate emotionally intelligent AI into our daily lives.
