Human vs. Artificial Intelligence: The Role of Agent Knowledge in Consumer Responses to AI Influencers, Moderated by Interactivity and Mediated by Anthropomorphism
通过两个实验，研究了消费者对算法意识的了解程度（人类智能）与AI影响者的交互性（人工智能）如何共同影响消费者态度和购买意愿，并发现拟人化是其中的中介机制。
This study conceptualizes the interaction between consumers and artificial intelligence (AI) influencers as a dynamic persuasion game between human intelligence and artificial intelligence. Specifically, we investigate how consumers’ agent knowledge of algorithmic awareness (representing human intelligence) and the interactivity of AI influencers (representing artificial intelligence) jointly shape consumer responses grounded in the theoretical frameworks of the persuasion knowledge model and the three-factor theory of anthropomorphism. Using a single-factor (algorithmic awareness: high vs. low) between-subjects experimental design, Study 1 demonstrated that consumers with high (vs. low) levels of algorithmic awareness exhibited significantly less favorable attitudes toward and purchase intentions for AI recommendation. Study 2, which employed a 2 (algorithmic awareness: high vs. low) × 2 (interactivity: high vs. low) between-subjects design, revealed that interactivity moderated the effect of algorithmic awareness. Specifically, the negative effect of algorithmic awareness on consumer responses emerged only in the high interactivity condition, but not in the low interactivity condition. Across both studies, anthropomorphism emerged as a mediating mechanism underlying these effects. The findings offer valuable theoretical insights and practical implications for understanding consumer behavior in AI-driven influencer marketing contexts.