Analyzing Human Decision Making within the Context of AI Interactions

Artificial intelligence (AI) is increasingly embedded in everyday decision-making, raising important questions about how reliance on AI shapes human cognition and behavior. This study analyzes the relationships between AI Dependence, Trust in AI, Decision Confidence, Cognitive Caution, and Independent Decision-Making using survey data from 201 respondents. Five hypotheses were tested using multiple regression, causal mediation analysis, and moderation analysis. Contrary to automation bias predictions, higher AI Dependence was associated with greater Cognitive Caution rather than reduced vigilance. Trust in AI was found to partially mediate the relationship between AI Dependence and Decision Confidence, accounting for approximately 24% of the total effect. AI Dependence was also associated with a modest but significant reduction in Independent Decision-Making. A moderation analysis testing whether Cognitive Caution buffers this reduction produced a non-significant interaction, though the pattern warrants further investigation. The findings challenge simplified accounts of automation bias and suggest that reliance and active cognitive engagement can coexist within AI-assisted decision-making.