Sounds good and feels better—how anthropomorphism influences artificial intelligence voice assistant use

Anthropomorphic design is increasingly embedded into artificial intelligence (AI) voice assistants (AIVAs) to create more emotionally engaging interaction experiences. However, despite widespread adoption, users continue to express significant concern regarding security-related risks, and the mechanisms through which anthropomorphic cues help alleviate this concern remain insufficiently understood. Drawing on dual-appraisal perspectives and affect-as-information theory, this study investigates how anthropomorphism can shape evaluative processing (i.e., perceived enjoyment, perceived security risk, and discontinuance intention) under conditions of technological opacity and uncertainty in AIVA use contexts. We develop a research model in which anthropomorphism functions not only as a generator of affect, but also as a contextual amplifier that strengthens the evaluative influence of affective experience during security-related judgment formation. Survey data from AIVA users were analyzed using structural equation modeling, supplemented by robustness checks and post-hoc analyses. The results show that anthropomorphism increases perceived enjoyment significantly, which in turn reduces both perceived security risk and discontinuance intention. However, anthropomorphism does not reduce perceived security risk directly; rather, the relationship operates primarily through perceived enjoyment, highlighting the central role of affective appraisal. Furthermore, anthropomorphism strengthens the negative relationship between perceived enjoyment and perceived security risk, suggesting that affective experience exerts a stronger evaluative influence under humanlike interaction conditions. Post hoc analyses further reveal nonlinear affective effects, highlight the importance of voice-specific anthropomorphic cues, and suggest that anthropomorphism may also facilitate habitual engagement through affective experience. Based on our findings, theoretical contributions, practical implications, and future research directions are provided.