Modeling Consumer Responses to AI-Generated Food Images: The Role of Visual Authenticity, Perceived Risk, and Trust

Abstract

The growing adoption of generative artificial intelligence (AI) in digital food marketing raises important questions regarding its impact on consumer behavior. This study aims to examine how AI-generated food images influence purchase intention by integrating perceived visual authenticity, perceived risk, and trust within the Stimulus–Organism–Response (S-O-R) framework. A quantitative explanatory approach was employed, using a cross-sectional survey of 204 Indonesian consumers who had purchased food online within the past six months. Data were analyzed using Structural Equation Modeling–Partial Least Squares (SEM-PLS). The findings reveal that AI-generated images significantly enhance both trust and purchase intention. Perceived visual authenticity emerges as the strongest predictor of trust and directly increases purchase intention, highlighting the importance of realistic, credible visual representations. In contrast, perceived risk does not directly affect purchase intention but negatively influences trust, indicating that its impact operates indirectly. Furthermore, trust plays a significant mediating role in the relationships between AI-generated images and purchase intention, as well as between perceived risk and purchase intention. However, trust does not mediate the relationship between perceived visual authenticity and purchase intention. This study concludes that consumer responses in AI-driven marketing contexts are primarily shaped by visual perception and trust rather than perceived risk. In practice, the findings suggest that businesses should prioritize authentic, transparent AI-generated visuals to strengthen consumer trust and enhance purchase intention.