Factors influencing single young adults’ intentions to accept AI virtual companion apps based on the theory of emotional design
Ting Liu, Yue Sun, Zheng-Qi Wei, Ting-Yun Lo · Frontiers in Psychology · 2026
Background AI virtual companion apps are increasingly used by single young adults as emotionally responsive media, yet acceptance studies often treat them as ordinary AI services or chatbots. This study examines how emotional design influences single young adults’ intention to accept AI virtual companion apps. Methods Drawing on emotional design theory, technology acceptance research, and human–AI companionship literature, a conceptual model was developed in which emotional design affects behavioral intention directly and indirectly through user experience. Emotional design was operationalized through four AI-companion-specific dimensions: interaction quality, visual appeal, perceived personalization, and intelligent adaptability. Measurement items were adapted from validated scales and refined through expert interviews. A survey was conducted with 635 Chinese single young adults who had used AI virtual companion apps for at least 1 month. Data were analyzed using reliability analysis, exploratory and confirmatory factor analyses, correlation analysis, structural equation modeling, and bootstrap mediation testing. Results The model showed adequate fit (CMIN/DF = 1.279; GFI = 0.980; AGFI = 0.971; IFI = 0.995; RMSEA = 0.021; CFI = 0.995; TLI = 0.994). Emotional design positively predicted user experience ( β = 0.447, p < 0.001) and behavioral intention ( β = 0.377, p < 0.001), while user experience also positively predicted behavioral intention ( β = 0.310, p < 0.001). Intelligent adaptability had the strongest correlation with acceptance intention ( r = 0.408, p < 0.01), followed by perceived personalization ( r = 0.371, p < 0.01). User experience partially mediated the relationship between emotional design and behavioral intention, with an indirect effect of 0.139 and a total effect of 0.515. Conclusion Intelligent adaptability and personalization appear especially important for the acceptance of AI virtual companion apps, while interaction quality and visual appeal support users’ initial and ongoing experiences. These findings clarify how AI-specific emotional design features become behaviorally meaningful through user experience and contribute to media psychology and human–AI interaction research.