Investigating consumer AI adoption in the banking sector: a cross-context meta-analysis of key determinants

Abstract

This study addresses the fragmented and inconsistent findings regarding consumer adoption of Artificial Intelligence (AI) in the banking sector. Through a meta-analysis of existing empirical research, it identifies and evaluates the most influential determinants of AI adoption. A meta-analysis was conducted using 58 empirical studies with data from 23,434 respondents across more than 20 countries. Seventeen drivers were examined, including traditional constructs (e.g., perceived usefulness, ease of use, trust, attitude) and post-adoption variables (e.g. satisfaction, customization, facilitating conditions). Moderation analysis assessed the influence of cultural and contextual variables using hierarchical meta-regression. Results confirm the central role of perceived usefulness, trust, attitude, and satisfaction in AI adoption, reinforcing the technology adoption models frameworks. Cultural and contextual moderators, such as Power Distance, Masculinity, Uncertainty Avoidance, Long-Term Orientation, and Human Development Index, significantly influenced these effects, highlighting the importance of sociocultural context. This research offers the first meta-analytic synthesis of consumer AI adoption in banking, integrating cognitive and post-adoption perspectives. Incorporating cultural and contextual moderators, it enhances theoretical generalization and provides actionable guidance for banks and policymakers implementing AI-based financial solutions.