Don’t Take it Personally: Resistance to Individually Targeted Recommendations with Anthropomorphic Recommender Systems

Abstract

Recommender agents, artificially intelligent recommender systems that demonstrate anthropomorphic cues, are widely available online to provide consumers with individually tailored recommendations. Nevertheless, little is known about the effect of their anthropomorphic cues on users’ resistance to both the system and recommendations. Moreover, individually tailored recommendations require users to disclose information proactively or reactively for receiving customized or personalized recommendations, which can trigger users’ resistance to the platform and the recommendations. Accordingly, this study examined the extent to which recommender systems’ anthropomorphic cues and the type of recommendations provided (customized and personalized) influenced online users’ perceptions of control, trustworthiness, and the risk of using the platform. The study assessed how these perceptions, in turn, influence users’ adherence to the recommendations. An online experiment among online users (N = 266) with recommender agents and web recommender platforms that provided customized or personalized restaurant recommendations was conducted. The results of the experiment entail that when recommendations are customized, as compared to personalized, users are less likely to demonstrate resistance and are more likely to adhere to the recommendations provided. Furthermore, the study’s findings suggest that these effects are amplified for recommender agents, demonstrating anthropomorphic cues, in contrast to traditional systems as web recommender platforms.