<![CDATA[Generative AI travel planners have become routine infrastructure, yet little remains known regarding which features Indonesian travelers actually experience as personalized and whether that experience confers emotional benefits. Drawing upon socio-technical systems theory and social exchange theory, this study investigates two technical elements (connectivity and ubiquity) and two social elements (dynamism and empathy) as antecedents of perceived personalization, while also testing inferred benevolent intentions (IBI) and inferred unselfish intentions (ISI) as dual social-exchange pathways. Data from an online survey of 180 Indonesian AI users were analyzed using PLS-SEM with 5,000 bootstrap resamples. The results demonstrate that connectivity (β = 0.416) and dynamism (β = 0.413) significantly predict perceived personalization, whereas ubiquity (β = 0.035) and empathy (β = 0.042) do not. Perceived personalization significantly increases both IBI (β = 0.518) and ISI (β = 0.445), and exerts a direct effect on hedonic well-being (β = 0.396). However, only the benevolent pathway is sustained: IBI significantly predicts hedonic well-being (β = 0.338), whereas ISI does not (β = -0.024). The model accounts for 39.8% of the variance in hedonic well-being. These findings affirm that personalization in AI travel planning originates from integrative and adaptive capabilities rather than mere availability or simulated warmth, and that Indonesian travelers convert personalization into felt enjoyment through perceived care rather than through perceived commercial neutrality.]]>
