Policy Simulation of Digital Leisure Adoption in Older Adults: A Psychosocial AI Cluster Approach

Abstract Background As digital transformation reshapes leisure and social participation, older adults remain disproportionately excluded from digital services. A previous study employing artificial neural networks (ANN) identified four psychosocial and behavioral clusters predicting adoption intent (Hwang et al., 2025). However, policy implications of these findings remain underexplored. This follow-up study applied a simulation-based framework to assess how tailored public health interventions could shift digital adoption intentions in segmented older adult groups. Methods Data from the 2022 Urban Policy Indicator Survey (N = 2,239; aged ≥50) were used. Variables included age, sex, education, income, household size, digital literacy, and adaptability. SPSS 23 was used for data cleaning and K-means clustering; ANN simulations were run using IBM SPSS Modeler 14.2. Four clusters were: • Cluster A: Digitally Empowered Midlife Men; • Cluster B: Digitally Marginalized Married Men; • Cluster C: Cohabitant Digital Pragmatists; • Cluster D: Socially Connected Digital Women. Policy scenarios were tailored to each group: AI/metaverse training (A), basic device provision with offline workshops (B), family-assisted mentoring (C), and social media-integrated wellness content (D). Simulations adjusted ANN input variables to estimate post-intervention adoption intent. Results Cluster B showed the highest increase in adoption intent (35.6%→62.4%, +26.8%), validating the effect of basic access and training. Cluster D rose from 58.2% to 80.3% (+22.1%) with tailored content. Cluster C increased moderately from 84.4% to 90.1% (+5.7%) with family-based support. Cluster A increased slightly (76.1%→78.9%, +2.8%), showing a need for content personalization. Conclusions Simulation-based policy modeling helps prioritize digital inclusion strategies. Tailored interventions significantly outperform generic approaches and improve digital engagement in aging populations. Key messages • AI-powered policy simulations provide actionable insights into enhancing digital inclusion among older adults. • Cluster-tailored public health strategies outperformed generic interventions in improving digital service adoption.