Yongxiang Gu, Yanhui Wang
As the development and application of Artificial Intelligence (AI) technologies accelerate, their disruptive impact on the labor market has become a central issue in economic research. This paper aims to investigate the multifaceted mechanisms through which AI influences income inequality. It begins by reviewing the established theoretical frameworks linking technological progress to income distribution, followed by a survey of contemporary literature on AI's labor market effects. Building on this foundation, the paper develops a theoretical framework that delineates four primary channels through which AI may exacerbate inequality: enhanced skill-biased technical change (SBTC), capital-labor substitution, job polarization, and the "winner-take-all" effect. The analysis suggests that while AI substitutes for routine tasks performed by low- and middle-skill workers, it simultaneously complements the cognitive, social, and creative skills of high-skill workers, thereby widening the wage gap. Furthermore, as a form of capital, AI deepens capital-labor substitution and may contribute to the rise of "superstar firms," potentially leading to a decline in the labor share of income and amplifying inequality between capital owners and labor. Based on these findings, the paper concludes by proposing a comprehensive suite of policy recommendations, including reforms in education, modernization of the social safety net, and adjustments in labor market regulation and technology governance. These policies are designed to mitigate the adverse distributional consequences of AI while harnessing its potential for productivity growth, ultimately fostering a more inclusive society.