Abstract
Artificial intelligence has evolved from rule-based systems to agentic AI: autonomous agents capable of perception, reasoning, and interaction with minimal human supervision. This transformation positions AI as a general-purpose technology with profound implications for labor markets. Autonomous agent capability is accelerating; the length of tasks frontier agents can complete unaided has been doubling roughly every seven months, yet measurement of labor effects has not kept pace, leaving managers and policymakers reliant on indices that capture only job elimination. This article reviews agentic AI through an interdisciplinary lens, integrating insights from computer science, economics, and organizational studies. It introduces the Labor Disruption Index (LDI), a composite framework capturing four dimensions of technological impact: displacement, automation potential, productivity enhancement, and new task creation. Unlike prior automation-risk indices that focus solely on job elimination, the LDI incorporates both negative and positive effects, enabling the identification of net outcomes at the occupational level. Each component is operationalized using existing datasets, and calibration procedures are demonstrated with sensitivity analyses across alternative weighting schemes. The analysis is organized through a micro–meso–macro framework: the micro level concerns agent design and technical capabilities; the meso level concerns organizational adoption patterns and governance structures; and the macro level concerns economy-wide impacts and policy responses. The synthesis draws on peer-reviewed empirical studies, including randomized controlled trials, adoption surveys, and macroeconomic projections. Findings reveal substantial heterogeneity across sectors: some roles face net displacement while others benefit from productivity gains and task creation. Applications are analyzed across healthcare, finance, manufacturing, legal services, education, and government. The article concludes with policy recommendations and an interdisciplinary research agenda organized around the LDI, advancing the measurement of labor disruption and providing actionable tools for scholars, managers, and policymakers.