Chatbots in Banking: Key Predictors of User Acceptance Across Two Large-Scale Studies—and How Gender and Ease of Use Fall Short

Conversational AI technologies are increasingly used in customer service settings, yet widespread adoption remains limited. Established models of technology acceptance offer valuable foundations but often overlook experiential and context-sensitive factors that shape customers’ willingness to use such systems. We present an analysis of two high-powered online studies (total N = 1,931) investigating predictors of intentions to use AI-based banking chatbots. Using a hierarchical regression model, we tested the predictive power of pragmatic (e.g., usefulness, ease of use), hedonic (e.g., enjoyment, relatedness), and person-specific (e.g., gender, openness to technology) factors. Our findings reveal that relevance, enjoyment, and usefulness are the strongest predictors of chatbot acceptance, while ease of use and gender show no explanatory power. We discuss potential reasons for the declining influence of these variables, which were considered fundamental to technology acceptance, and highlight the need to reorient design and evaluation efforts toward experiential dimensions of human-chatbot interaction.