A growing body of economic research explores the possible consequences of artificial intelligence replacing or augmenting human labor, yet little is known about how AI might affect worker power. This is despite the fact that the same aspects of AI that make it uniquely disruptive as a tool of automation—carrying out non-routine tasks, assimilating tacit knowledge, refining predictions—also position AI for applications in management and human resources. This scoping review surveys and reframes the economic literature on AI and labor, detailing how new technologies might alter worker power with consequences for income distribution and job quality. To help clarify the theoretical effects of AI on workers and the nature of work, I draw a distinction between the labor demand impacts and the worker power impacts of new technologies (or, between AI’s effects on job content versus job context). Labor demand impacts stem from AI that substitutes for tasks now carried out by humans. Worker power impacts stem from AI that sharpens managerial control, undermines workplace norms, or heightens information asymmetries. With this framework in mind, I formalize several avenues by which AI potentially shifts power dynamics in workplace and bargaining contexts, including monitoring and surveillance, predictive analytics in wage bargaining, and algorithmic management systems that diminish worker autonomy. This study links the evolving literature on the economics of AI to research strands concerned with the determinants of worker power and its long-term trends.


Davis, Owen F. (2024). “Artificial Intelligence and Worker Power.” Working paper.