Can AI Write Authentic CSR ? Effects of Message Source Across Informational and Narrative CSR Communication

ABSTRACT As artificial intelligence (AI)‐generated messages increasingly approach human‐level quality, can AI effectively substitute human authors in value‐laden brand communications such as corporate social responsibility (CSR) messaging? This study examines how CSR message source (AI‐generated vs. human‐written) influences consumer perceptions of message authenticity and credibility and subsequently shapes brand attitudes and CSR support intentions. Two experiments test the source effects across two message formats: informational and narrative. Results indicate that human‐written messages consistently generate higher authenticity and more favorable CSR outcomes. Credibility emerges as a significant mediator in informational messages, while machine heuristic serves as a boundary condition in narrative messages—suggesting that evaluations of AI‐generated CSR messages depend on the message format. This study contributes to CSR literature by identifying the format‐specific evaluative mechanisms that shape consumer responses to AI‐generated CSR communication and offers practical implications for the strategic use of AI to enhance authenticity and credibility in CSR communication.