Research Papers
Gu Chenyan, Chang Xinsiyu, Lai Jiaying, Jia Shuyue, Zhou Lei
2026, 6(1): 134-152.
Abstract With the continuous iteration and innovation of generative AI, the realism of AI-generated model ads is increasingly approaching that of ads featuring real people. AI can even transcend reality, creating models with hyper-realistic physical characteristics. Exploring consumer responses to these types of ads is becoming an important issue. Based on existing literature on AI-generated ads, this study focuses on the effects and mechanisms of models with different degrees of authenticity in AI-generated ads. Through five experiments, this study reveals that the higher authenticity of models in AI-generated ads, the more positive the consumer responses. The mediating effect further indicates that the higher authenticity, the lower eeriness, which in turn leads to more positive advertising attitudes and purchase intentions. The above effect is moderated by AI disclosure, brand type and brand popularity. Specifically, when the brand discloses the AI generation, or when the brand has high popularity, the advantage of lifelike image is significantly reduced; for innovative brand, the advantage is notably enhanced. This study enriches the theoretical research on AI-generated ads and offers valuable management insights for brands to conduct advertising.