Out of gratification or fears? A dual model to explore drivers of generative artificial intelligence adoption

Generative artificial intelligence (GenAI) has emerged as a transformative force reshaping how individuals work, live, and interact with their environments. Despite its rapid diffusion, research has yet to clarify the psychological mechanisms driving individual-level GenAI adoption and resistance. This study addresses this critical gap by proposing a dual-path model grounded in Uses and Gratifications Theory (UGT) and an extended Protection Motivation Theory (PMT) framework. UGT explains the positive, need-fulfilling motivations for GenAI use, incorporating novel gratification constructs such as serendipity, perceived diagnosticity, tangibility, curiosity fulfillment, and enjoyment. In contrast, the enhanced PMT framework captures both traditional and extended pathways of perceived threats. By including fear of losing power (FLP) and fear of missing out (FoMO) as internal psychological mechanisms, this study offers a more comprehensive account of GenAI adoption and resistance. Trait competitiveness and AI self-efficacy are introduced as moderators, delineating how individual differences shape protective responses. Using data from a two-wave longitudinal survey of 1271 ChatGPT users, the findings reveal that UGT-related factors primarily drive adoption, while traditional and extended PMT factors explain resistance behaviors. Notably, FoMO functions as a dual-pathway factor, facilitating adoption and mitigating resistance. Trait competitiveness and AI self-efficacy demonstrate partial moderating effects, underscoring the role of personal dispositions in shaping user behavior. This study contributes theoretically by integrating positive gratification and protective aversion into a unified model of GenAI use. Practically, it provides actionable insights for designing adaptive, user-centered AI systems that enhance engagement while reducing resistance.