Understanding the influence of AI labels on consumer psychology: the moderating role of product type

Driven by rapid progress in generative artificial intelligence, the use of AI-generated models in online product displays has become increasingly widespread. Although AI-generated content appears highly realistic, it is commonly associated with adverse consumer responses. This study examines how labeling a model in a product image as AI-generated (vs. human) influences consumers' self-expression. The findings show that labeling a model as AI-generated functions as an emphasis-framing cue that elevates consumers' sense of eeriness, which subsequently increases both perceived psychological risk and perceived performance risk, thereby reducing consumers' self-expressive tendencies. This process is moderated by product type, such that the negative effect is amplified for symbolic products, whereas it disappears for functional products. Building on theories of framing effect, algorithm aversion, emotion, and risk perception, this study elucidates how AI-generated labels shape consumers' self-expression and identifies the boundary conditions of these effects across different product types.