Intolerance of Uncertainty and Artificial Intelligence Dependency: The Mediating Role of Metacognitive Confidence among Remote and In-Office IT Startup Employees

Abstract

This study examined the mediating role of metacognitive confidence in the relationship between intolerance of uncertainty and AI dependency among IT startup employees. This is particularly relevant given that approximately 72% of employees report daily AI usage, with remote workers showing heightened cognitive reliance on AI tools. Data were collected from 372 Indian IT startup employees (remote n = 192, in-office n = 180) using the IUS-12, MCQ-30 Cognitive Confidence subscale, and DAI-5. Data were analyzed using descriptive statistics, Pearson correlations, independent samples t-tests, and mediation analysis via PROCESS Model 4 in JAMOVI. Results indicated that intolerance of uncertainty significantly reduced metacognitive confidence (b = −0.117, p < .001), which in turn increased AI dependency (indirect effect b = 0.0163, p = .003, 95% CI [0.0066, 0.0275]), alongside a significant direct effect (b = 0.201, p < .001). Remote workers had lower confidence (t=-2.274, p=.024) and higher dependency (t=3.459, p=.001) compared to in-office employees; females showed higher metacognitive confidence (t=2.163, p=.031).  These findings explain how uncertainty erodes cognitive self-trust and increases AI dependence among IT startup employees, highlighting the importance of maintaining cognitive autonomy in hybrid technology environments.