Modeling public trust in AI cognitive capabilities using statistical and machine learning approaches

Abstract

Abstract As artificial intelligence (AI) systems increasingly perform cognitive functions, assessing public trust in these capabilities is critical. This study investigates the impact of age, gender, and familiarity with AI on confidence in AI’s ability to make simple decisions, complex judgments, and perform memory recall tasks. A survey of 400 participants was analyzed using statistical tests and a Random Forest classifier. Results indicate that AI familiarity is the strongest predictor of confidence, followed by age and gender. Participants expressed greater trust in AI for factual, memory-based tasks, and preferred human decision-making in high-stakes scenarios such as medical diagnosis and autonomous driving. The Random Forest model demonstrated strong predictive performance, confirming that familiarity and age are the most influential predictors of trust. These findings highlight the nuanced role of demographic and experiential factors in shaping trust in AI’s cognitive capabilities and provide practical implications for designing user-aligned, trustworthy AI systems.