The Role of AI Streamers in Live-Streaming E-Commerce: Examining the Impact of Interaction on Consumer Trust and Purchase Intention

As AI-driven technologies become increasingly integrated into digital commerce, understanding how consumers perceive and interact with AI streamers in live shopping environments has become crucial. This study investigates how varying levels of AI streamer interactivity influence consumer trust and purchase intention, with perceived usefulness (PU) and perceived ease of use (PEU) as key mediating variables. Grounded in the Technology Acceptance Model (TAM), a controlled between-subjects experiment was conducted with 157 Chinese-speaking participants, who were randomly assigned to one of three interactivity conditions: low, medium, or high. Data were analyzed using regression analysis and mediation testing via the Model 4 of Hayes' (2022) PROCESS macro for SPSS (Version 4.2). The results showed that higher AI interactivity significantly enhanced both PU and PEU, which subsequently increased trust in the AI streamer. Further mediation analyses confirmed that PU and PEU partially mediated the relationship between AI interactivity and trust. Additionally, consumer trust strongly predicted purchase intention. These findings contribute to the growing body of literature on AI-human interaction in digital commerce and provide actionable insights for e-commerce platforms aiming to optimize AI streamer design to build consumer trust and drive purchasing behavior.