AI dependency and its relationships with anxiety, quality of life, and digital stress

Abstract

The integration of artificial intelligence (AI) into higher education has accelerated, yet little is known about the psychological mechanisms underlying students’ reliance on AI. This study conceptualizes AI dependency as a complex cognitive–motivational construct that extends beyond mere usage, influencing anxiety, digital stress, and quality of life. A total of 521 participants (predominantly undergraduate, 80% female) were recruited via snowball sampling at King Abdulaziz University. Self-administered standardized instruments assessed AI dependency, AI-related general anxiety, digital stress, and quality of life. A network analysis approach was employed to examine the interrelations among AI dependency, cognitive offloading, anxiety, availability pressure, FoMO, digital overload, digital vigilance, social acceptance anxiety, and quality of life among university students. This approach allowed identification of central variables and conditional interactions within a dynamic psychological system. AI dependency emerged as a central node in the network, showing strong conditional associations with various digital stressors, including digital vigilance and fear of missing out. Anxiety appeared to occupy a bridging position in the network, linking cognitive reliance to environmental pressures, while these patterns of association were, in turn, related to reduced quality of life. Social acceptance anxiety translated cognitive pressures into relational–identity concerns, and cumulative effects manifested in reduced quality of life. The network revealed non-linear, conditional associations, highlighting that the psychological impact of AI dependency is mediated by cognitive, motivational, and contextual factors rather than by direct usage intensity alone. AI dependency is not a neutral or purely functional behavior but a central psychological construct with both potential advantages, such as reduced cognitive load and increased efficiency, and risks, including diminished autonomy, heightened anxiety, and long-term digital strain. These findings offer a culturally contextualized model for understanding AI’s influence on student well-being and provide a framework for interventions that target central nodes in the network to promote healthier engagement with AI in academic settings.