Consumer Autonomy in Generative AI Services: The Role of Task Difficulty and AI Design Elements in Enhancing Trust, Satisfaction, and Usage Intention

Abstract

As generative AI services become increasingly integrated into consumer decision making, concerns have grown regarding their influence on consumer autonomy—the extent to which individuals retain independent control over AI-assisted decisions. Although these services offer efficiency and convenience, they can simultaneously constrain consumer decision making, potentially impacting trust, satisfaction, and usage intention. This study investigates the role of perceived consumer autonomy in shaping consumer responses, specifically examining how task difficulty (Study 1) and AI service design elements—explainability, feedback, and shared responsibility (Study 2)—influence autonomy perceptions and subsequent consumer outcomes. Using two scenario-based experiments involving a total of 708 participants, the results reveal that perceived autonomy significantly enhances consumer trust, particularly in contexts involving high task difficulty. Among the tested AI design interventions, shared responsibility emerged as most effective in enhancing perceived autonomy, trust, satisfaction, and long-term engagement, whereas explainability and feedback alone showed limited impact. These findings underscore the importance of designing AI services that actively support consumer agency through user-involved decision-making frameworks rather than relying solely on passive informational transparency. Theoretical implications for consumer autonomy in AI interactions are discussed, along with practical recommendations for designing consumer-centered AI services.