The Impact of Generative AI Images on Consumer Attitudes in Advertising

作者：Lei Zhang, Chung Hur · 发表于：Administrative Sciences · 年份：2025 · DOI：10.3390/admsci15100395 · 被引用次数：19 · 研究领域：AI in Service Interactions、Impact of AI and Big Data on Business and Society、Consumer Behavior in Brand Consumption and Identification

While the capability of generative AI to generate high-quality content is well-recognized, there is still a lack of in-depth research on its actual impact on marketing effectiveness within real-world marketing environments. This study addresses this gap by conducting experiments to examine the effects of AI-generated advertisement images, created using text-to-image diffusion models, on consumer responses and the boundary conditions of these effects. Study 1 (n = 130) found that for coffee ads, attitudes were descriptively higher toward AI-generated images (ηp2 = 0.17), whereas for medical-aesthetics and public-service ads, evaluations favored human-made images; none of these differences reached significance. Study 2 (n = 79) revealed that when consumers were informed about the source of the image (AI or human), they showed significantly more positive attitudes toward human-made images than those generated by AI (d = 0.52). Study 3 (n = 209) demonstrated that in commercial advertising contexts where usage motivations were disclosed, consumers’ negative reactions to AI-generated images were moderated by the specific usage motivation (η2 = 0.04). When the motivation was privacy protection, evaluations were comparable to human-made images. In contrast, visual appeal produced slightly lower but non-significant ratings, whereas cost efficiency led to significant declines in trust and purchase intention, with attitude showing only marginal decreases and preference no differences. T...