Afraid of AI, Addicted to AI: Unraveling the Fear–Dependency Paradox
Author:
Open Access
- 1 July 2026
- journal article
- research article
- Published by SAGE Publications in Sage Open
- Vol. 16 (3)
Abstract
Classical technology acceptance theory treats fear as a barrier to adoption. Yet emerging evidence suggests that in professional AI contexts, fear may paradoxically drive rather than deter usage. This study investigates this counterintuitive dynamic by examining the relationship between AI anxiety and AI dependency among 421 working adults in Turkey, using structural equation modeling (SEM) and multi-group analysis (MGA). We introduce the “control paradox”—a novel theoretical construct proposing that individuals who fear losing cognitive autonomy to AI increasingly surrender that autonomy through defensive, compulsive overuse. Four principal findings emerge. First, AI anxiety comprises four empirically validated dimensions—learning anxiety, job displacement fear, sociotechnical blindness, and AI configuration anxiety—with distinct effects on dependency. Second, job displacement fear is the strongest predictor of dependency (β = 0.474, p < .001), revealing a professional threat-response mechanism not captured by existing TAM/UTAUT frameworks. Third, sociotechnical awareness functions as a protective factor (β = −0.225, p < .01), demonstrating that critical AI literacy—rather than technical skill alone—disrupts the fear–dependency cycle. Fourth, MGA reveals that the job displacement fear–dependency relationship is gender-invariant (males: β = 0.360, p < .001; females: β = 0.329, p = .006; Fisher’s Z = 0.285, p = .776), establishing the control paradox as a structural feature of professional AI adoption that transcends demographic variation. These findings extend technology acceptance theory, reframe AI dependency as professionally motivated rather than deficit-driven, and yield concrete recommendations for AI literacy education, workforce policy, and ethical technology design.Keywords
This publication has 68 references indexed in Scilit:
- Investigating User Acceptance of AI-Driven Fintech Services Through the Lens of the UTAUT Model2025
- WEIRD? Institutions and consumers’ perceptions of artificial intelligence in 31 countriesAI & SOCIETY, 2025
- Risk and artificial general intelligenceAI & SOCIETY, 2024
- One is the loneliest number… Two can be as bad as one. The influence of AI Friendship Apps on users' well‐being and addictionPsychology & Marketing, 2023
- What Makes AI Addictive? The Role of Discounting, Risk Aversion and Self-regulation2023
- Negative Social Impacts of Artificial Intelligence and the Main Mitigation Actions: A Systematic Review2023
- AI as a boss? A national US survey of predispositions governing comfort with expanded AI roles in societyAI & SOCIETY, 2021
- The Art and Science of User Exploitation: AI in the UAE and Beyond2021
- Ethical, Societal, Philosophical, and Spiritual Issues2020
- An Index of Factorial SimplicityPsychometrika, 1974