The study explores how the use of generative artificial intelligence (GenAI) affects donor empathy in medical crowdfunding. It draws on the empathy-altruism hypothesis and tests whether GenAI amplifies empathic engagement and induces altruism. A pre-registered between-subjects experiment (N = 601) was conducted, featuring a 2 (Authorship: Human, GenAI) x 2 (Disclosure Status: Disclosed, Hidden) factorial design. Results suggest that GenAI is comparable to human authors in its ability to elicit empathy, which subsequently increases altruistic intentions and donation amount. However, when donors become aware of GenAI’s involvement in narrative writing, empathy becomes significantly suppressed that eventually inhibits donation intentions and behavior. The finding that GenAI transparency is costly offers new insights into the theories of algorithm aversion and artificial empathy. It highlights the neglected economic consequences and moral implications associated with GenAI disclosure, which is increasingly mandated across different domains.