Understanding Continuance Intention toward AI Virtual Streamers in E-Commerce: An Expectation Confirmation Perspective

With the rapid growth of AI virtual streamers in E-commerce, understanding what drives viewers’ continuance intention has become increasingly important. Although AI virtual streamers are increasingly adopted by platforms and brands, existing research has not sufficiently explained how viewers form expectations toward these AI agents, how these expectations are confirmed through actual viewing experiences, and why viewers choose to continue watching them. Addressing this gap, this study draws on expectation confirmation theory to examine how three key characteristics of AI virtual streamers, anthropomorphism, emotional richness, and interactivity influence viewers’ satisfaction and, in turn, their continuance intention. To deepen the analysis, the study introduces AI literacy as a moderating factor, proposing that viewers with different levels of AI understanding may evaluate AI streamer features differently and therefore experience varying levels of satisfaction. By integrating evaluation processes with AI-specific user differences, this study provides a more comprehensive explanation of continuance intention in AI-mediated live streaming commerce. The findings aim to advance theoretical discussions on AI influencers, user expectations, and technology acceptance, while offering practical insights for platforms and brands seeking to design AI virtual streamers that foster sustained viewer engagement and create long-term competitive advantages in the live-streaming commerce market.