AI-Generated Human Models in Advertising: Effects of Product Type and Message Appeal on Mind Perception, Perceived Fit, Message Credibility, and Ad Attitude

Brands are increasingly deploying artificial intelligence (AI)-generated human models in advertising as substitutes for traditional human models. Drawing on Mind Perception Theory, we propose that consumers attribute high agency but low experience to AI human models and that this asymmetric source profile produces differential advertising effectiveness across product categories. Across three studies (Study 1: N = 96; Study 2: N = 142; Study 3: N = 347), we examine the following: (1) whether AI-human models are more effective for utilitarian than for hedonic products; (2) the source-perception mechanism that explains this differential effectiveness; and (3) whether message appeal moderates the mechanism. Results show that AI-generated humans in advertisements are more favorably evaluated for utilitarian products, an effect that is serially mediated by perceived agency, perceived source–product fit, and message credibility. Perceived experience does not mediate. A product-by-message-appeal interaction further emerges in Study 3, demonstrating a match-up effect in which utilitarian-informational and hedonic-emotional pairings produce the most favorable ad attitudes. The agency-driven serial mediation operates as a parallel mechanism that is not amplified by message-appeal congruence. Theoretical and managerial implications are also discussed.