Does Source Credibility Theory Apply to Virtual Influencers? Evidence From Italy

The rapid evolution of Generative Artificial Intelligence has led to the emergence of increasingly realistic human-like virtual influencers (VIs), revolutionizing influencer marketing. While previous literature has focused primarily on clearly artificial or stylized digital figures, little is known about how consumers perceive the new generation of hyper-realistic VIs, nearly indistinguishable from human influencers. In particular, it is still unclear whether Source Credibility Theory (SCT), widely validated in the context of human endorsements, is also applicable when the source is artificially generated. This study aims to test the applicability of SCT in the context of human-like VIs, analyzing to what extent the three central dimensions of credibility (attractiveness, trustworthiness, expertise) influence consumers' attitudes towards these artificial endorsers. The research is based on a questionnaire administered to 372 Italian respondents and combines descriptive analysis and a multiple linear regression to answer three research questions: (1) how human-like VIs are perceived by Italian consumers; (2) whether they are considered as attractive, expert and trustworthy as human influencers; (3) to what extent the dimensions of credibility influence attitudes towards them. The results show that attitudes toward VIs are relatively low, despite the three credibility components having moderate values. VIs are perceived as quite attractive and expert compared to human influencers, but slightly less trustworthy. Furthermore, the regression analysis highlights that all three credibility dimensions positively and significantly influence attitudes, with attractiveness emerging as the strongest predictor. This confirms that the classic mechanisms of SCT remain valid even when the source is not human but artificially generated. The study contributes to the literature by providing empirical evidence on the perception of human-like VIs in the Italian context and testing the robustness of a well-established theory in influencer marketing in the AI context. Managerial implications are also discussed to guide how brands can strategically implement human-like VIs within the Italian market, which appears not yet fully ready to embrace these artificial endorsers.