Beyond the prompt: exploring technostress, flow and personality in generative AI behaviors

Purpose This study aims to investigate the dual psychological responses of users toward generative AI tools, focusing on technostress as a critical stimulus and examining its impact on user flow, continuance intention (CI) and switching intention (SI). It also explores the moderating role of autotelic personality (AP) to understand individual differences in coping with generative AI-induced demands. Design/methodology/approach Integrating the stimulus-organism-response (SOR) model and flow theory, a three-wave time-lagged survey design was used to mitigate common method bias and capture temporal dynamics in user behavior. Data were collected from 333 valid respondents across three time points. Findings The results reveal that technostress reduces flow experience and CI while increasing SI. AP significantly moderates these relationships, such that individuals with high autotelic traits demonstrate psychological resilience, maintaining flow and continuance while resisting switching, even under high technostress. Practical implications The findings yield several valuable practical insights for GenAI developers and digital library designers who integrate GenAI in information services, offering actionable strategies to enhance user engagement, reduce technostress and promote sustainable adoption in information-rich contexts. Originality/value By embedding flow theory within the SOR framework, this study offers a novel theoretical lens to explain users’ emotional ambivalence in AI-mediated environments. It contributes to emerging scholarship on technostress, intrinsic motivation and post-adoption behavior, responding to recent calls in the Electronic Library for understanding GenAI’s broader implications on digital engagement.