How do AI micro-dramas lead to addictive use among older adults? A chain mediation model of escapism motivation and flow experience.

Generative AI is deeply involved in the production and recommendation distribution of micro-short dramas, allowing older adult users to face a higher risk of out-of-control use while obtaining emotional compensation and immersive experience. This study mainly examines how three types of characteristics of AI micro-drama content (emotional realism, narrative involvement, and algorithmic personalization) affect the addictive use of older adults, and examines the chain mediation model between escapism motivation and flow experience. A total of 658 valid samples were obtained by questionnaire survey. SmartPLS 4 was used for data analysis to test the measurement and structural model, and the direct and indirect effects were estimated by Bootstrap. The results show that the three types of characteristics can not only directly and positively predict addictive use, but also significantly improve escapism motivation and flow experience; Escapism motivation and flow experience further positively predict addictive use, and the role of central flow is more prominent. The mediation test supports multiple single mediation and sequential mediation paths of "escapism motivation → flow experience," revealing that older adults are more likely to experience a gradual process from "entering for escape" to "difficult to quit after immersion." The research provides actionable governance targets for the design of age-friendly platforms and digital health interventions in the AIGC era.