24th MEB Conference - Budapest, 24.04.2026

With the advancement of 5G network technology and artificial intelligence, live streaming has become one of the increasingly important channels in the e-commerce field, including not only live streaming by human streamers, but also live streaming by AI streamers.This study compares the dynamics of AI-powered and human streamers in live-streaming e-commerce, with an emphasis on audience retention patterns, fatigue resilience, and algorithmic interactions. The study presents a multi-dimensional approach to assessing streamers' endurance and viewer engagement sustainability. The study examines at how the operational stability of AI streamers and the adaptive capabilities of human streamers differently affect long-tail audience retention and algorithmic visibility through observational data capturing real-time viewership metrics and platform-level interventions. The framework presents the idea of "algorithmic acclimation" to quantify platform-driven traffic compensation mechanisms triggered by consistent performance metrics. Preliminary results reveal that AI and human streamers have different resilience characteristics, which has consequences for content strategy optimization. By developing a fatigue-inclusive evaluation model, this study advances the understanding of human-AI coexistence in digital retail ecosystems while also providing organizations with strategic insights for streamers deployment.