Individual Privacy Perception in the Digital Age: The Interaction of Artificial Intelligence Attitude and Dependency

The increasing dependence on AI-supported services raises important questions about how positive beliefs about AI can turn into privacy risks. This study tests a gender-moderated mediation model of AI attitude, AI dependence, and online privacy concern (OPC) among Turkish university students. A cross-sectional survey conducted on 478 students using validated scales (AIAS-4, AI Dependency Scale, OPC Scale) was analyzed using structural equation modeling and the PROCESS Model 59. The measurement model demonstrated excellent fit (χ²/df = 1.01, CFI = 0.999, RMSEA = 0.005) and strong reliability-validity indicators. AI attitude significantly increased AI dependency (β = .50, p < .001), which in turn strengthened OPC (β = .77, p < .001). Gender moderates both relationships and reveals a significant moderator-mediation index (−.11; 95% CI [−.21, −.01]). Overall, the model explains 28% of the variance in OPC. The findings reveal a two-way effect of positive AI attitudes: while promoting beneficial participation, they also increase dependency-based privacy concerns, particularly among female users. Organizations should integrate privacy-aware AI literacy and gender-sensitive feedback mechanisms into digital platforms to mitigate risks while maintaining trust.