This study compares the effects of AI and human broadcasters on consumer engagement, trust, satisfaction, and impulsive consumption in live-streaming e-commerce. Drawing on the Stimulus–Organism–Response (S-O-R) framework, broadcaster type is conceptualised as the stimulus, perceived status elevation and Consumer senses are examined as organismic states, and impulsive consumption is treated as the behavioural response. Social overload is incorporated as a moderating condition. A quantitative survey design was adopted, and data were collected from consumers with experience in AI- and human-led live-streaming commerce. The proposed model was tested using PLS-SEM and multi-group analysis to examine direct, mediating, and moderating effects. The results show that both AI and human broadcasters positively influence perceived status elevation and Consumer senses. Human broadcasters exert a stronger effect on Consumer senses, whereas AI broadcasters provide more structured and consistent interaction cues. Social overload positively moderates the relationship between AI broadcasters and Consumer senses. Social overload also positively moderates the relationship between human broadcasters and Consumer senses. In addition, social overload moderates the relationship between Consumer senses and consumer engagement only in the human broadcaster group. Trust and consumer satisfaction significantly predict impulsive consumption in both groups, while consumer engagement predicts impulsive consumption only in the human broadcaster group. This study contributes to live-streaming commerce research by directly comparing AI and human broadcasters within an integrated mediation–moderation framework. It clarifies how broadcaster type shapes consumer responses through perceived status elevation, Consumer senses, satisfaction, engagement, and trust, while also showing the context-dependent role of social overload.

