The Illusion of Care: Personalization Intensity and the Authenticity Paradox in Human and AI-Mediated Hotel Service Encounters

Abstract

This study examines the personalization–authenticity paradox in human and AI-mediated hotel service encounters, proposing that personalization enhances perceived relational authenticity up to an optimal level but becomes counterproductive when perceived as excessive or calculative, while also examining whether the curvature of this relationship differs between human and AI-mediated service providers and whether authenticity transmits effects to emotional trust and loyalty intentions. A 3 × 3 between-subjects experiment crossed service-agent type (human, disclosed AI, source-ambiguous AI) with personalization intensity (low, moderate, high) among 270 hotel guests, with data analyzed using factorial ANOVA, planned polynomial contrasts, and bootstrap conditional-process analyses. Perceived authenticity exhibited a significant inverted-U relationship with personalization intensity, with authenticity highest under moderate personalization. Human service generated higher authenticity than disclosed-AI service, which in turn generated higher authenticity than source-ambiguous AI service. The omnibus Agent × Personalization interaction was not significant, and the theoretically focal Agent × Quadratic interaction was also insignificant, providing no statistical evidence that agent type altered the curvature of the personalization–authenticity relationship. Perceived authenticity positively predicted emotional trust and loyalty intentions and mediated the relationship between personalization curvature and both outcomes, while the negative quadratic association was stronger at higher levels of service criticality. The findings shift the personalization debate from whether hotels should personalize to how personalization should be calibrated, proposing a personalization meaning transformation from recognition, to understanding, to perceived calculation as a theoretical explanation for why excessive personalization may undermine authenticity across human and AI-mediated service encounters.