A systems view of consumer trust and engagement with AI-powered virtual influencers in online shopping

Abstract

This study aims to investigate the key enablers associated with consumer responses towards artificial intelligence (AI)-powered virtual influencers (VIs) in online shopping, with a focus on the psychological, technological, communication-related and cultural factors shaping consumer trust, engagement and decision-making. A multi-phase mixed-method design was used. First, 14 enablers were identified through a focused literature review and refined through expert validation involving 46 academic and industry professionals. Second, a Consumer-Informed Interpretive Structural Modelling (ISM) and Matrice d’Impacts Croisés Multiplication Appliquée à un Classement (MICMAC) analysis were conducted using pairwise influence judgements obtained from 181 consumers. The resulting framework was used to examine hierarchical relationships among the enablers and classify them according to their driving and dependence power. The ISM results reveal a nine-level hierarchy in which technological sophistication, platform compatibility and cultural relevance function as foundational driving enablers. Perceived trustworthiness, perceived authenticity and parasocial interaction emerge as intermediate relational mechanisms, while engagement quality, emotional appeal and social proof represent the most dependent outcome-orientated enablers. MICMAC analysis further classifies the enablers into independent, autonomous and dependent clusters, demonstrating how technological, cultural and communication-related factors collectively shape consumer evaluations of AI-driven VIs. The resulting hierarchy represents consumers’ perceived cognitive structure of virtual influencer enablers within online shopping environments. Drawing on the technology acceptance model, elaboration likelihood model, source credibility theory and persuasion knowledge theory as interpretive lenses, the study provides a systems-level understanding of the interrelationships among virtual influencer enablers. While limited by its consumer-informed structural modelling approach and its focus on an urban Indian context, the framework offers a foundation for future cross-cultural, longitudinal and explanatory research. The findings provide marketers and platform managers with a structured roadmap for designing culturally relevant, technologically robust and trustworthy virtual influencer strategies. The results further highlight the importance of transparency, responsible AI disclosure, consumer literacy and ethical governance in AI-mediated marketing environments. To the best of the authors’ knowledge, this study is among the first to apply a Consumer-Informed ISM–MICMAC approach to virtual influencer marketing, offering a novel systems-based perspective on the hierarchical relationships among enablers associated with consumer responses in AI-driven digital commerce.