AI Adoption in Livestreaming Supply Chains: How to Outperform Humans?

The rapid diffusion of artificial intelligence (AI) technologies has given rise to AI-livestreaming in online retail, raising a fundamental question of whether and when AI can replace human streamers in livestreaming e-commerce. This paper develops analytical models to examine a retailer’s strategic choice among no-livestreaming, human-livestreaming, and AI-livestreaming. We consider livestreaming into a Hotelling-type consumer mismatch framework in which human-livestreaming reduces perceived mismatch through information enhancement, while AI-livestreaming combines imperfect imitation of human persuasive ability with responsiveness-based value creation under convex technology costs. Our analysis characterizes equilibrium prices, demand, and profits under each strategy and reveals informative insights. While both human- and AI-livestreaming improve pricing power relative to no-livestreaming, their profit and demand implications differ markedly. Human-livestreaming may reduce profitability or fail to expand demand when consumer valuation is moderate, whereas AI-livestreaming generates more robust demand expansion and can strictly dominate under feasible technology conditions. The retailer’s optimal technology choice is governed by threshold effects in imitation capability, mismatch intensity, and cost parameters. Extending the analysis to a competitive setting further shows that allowing AI responsiveness to adjust endogenously substantially enlarges the region in which AI-livestreaming achieves simultaneous price and profit dominance.