Artificial Intelligence and the Restructuring of Saudi Labor Markets: A Statistical Analysis of Job Displacement and Skill Transformation

Abstract

Abstract The integration of artificial intelligence into state-led rentier economies presents unique structural challenges, particularly within Saudi Arabia’s Vision 2030 framework. Utilizing a balanced panel dataset of 120 sector-year observations across 10 economic sectors over a 12-year period, this study examines the structural association between AI adoption and the employment share of Saudi nationals. To address prevalent limitations in composite index construction and prevent methodological overfitting, the analysis employs a parsimonious econometric strategy. The AI Adoption Index is rigorously validated through internal consistency testing (Cronbach’s α = 0.87) and PCA-based sensitivity analysis, ensuring robustness beyond arbitrary weighting conventions. The empirical framework relies on fixed-effects panel regression supplemented by an instrumental variable (IV) approach to mitigate potential endogeneity, alongside a comprehensive battery of diagnostic tests (Hausman, Modified Wald, and Wooldridge). The findings reveal a statistically significant negative association between AI adoption and national employment in routine-intensive sectors (β = −0.41, p < 0.05), contrasted by a positive correlation with high-skill digital competencies (β = +0.67, p < 0.01). Theoretically, these structural shifts are interpreted through a macro-sociological lens of digital labor segmentation, explicitly distinguishing between observable macro-level reconfigurations and micro-level subjective experiences without claiming to measure individual worker consciousness. The study concludes with targeted policy recommendations for inclusive AI governance and sector-specific upskilling, ensuring digital transition strategies align with the social equity objectives of the state.