An Exploration of AI Adoption and Public Sector Operations: The Mediating Effect of Strategic Decision Making

Abstract

Purpose: While artificial intelligence (AI) offers the potential to enhance efficiency, accuracy, and responsiveness, empirical evidence suggests that its direct impact on public sector operations remains limited. This raises important questions about the mechanisms through which AI generates value in public administration. Methodology: This study employed primary data because primary data collection ensures control over data quality and relevance, which is crucial for testing the proposed hypotheses and generating actionable insights for government decision-makers. Quota sampling, which is a non-probability sampling technique that ensures representation from specific subgroups within a population, was used to obtain a sample size of 545. Findings: The AI Adoption and Strategic Decision-Making path shows a strong, positive, and statistically significant effect of AI adoption on strategic decision-making. The Strategic Decision-Making and Public Sector Operations path also demonstrates a moderate-to-strong positive effect that is statistically significant. The direct path from AI adoption to public sector operations is very weak and statistically insignificant. Strategic decision-making fully mediates the relationship between AI adoption and public sector operations. Unique Contribution to Theory, Policy, and Practice: The finding that strategic decision-making fully mediates the relationship between AI adoption and public sector operations adds theoretical depth by introducing a mechanistic pathway. This addresses a notable gap in the literature where most studies assume a direct link between digital tools and performance. The research cautions against policy approaches that focus solely on technology procurement. Instead, it encourages a shift toward policies that promote AI-supported strategic planning frameworks, ensuring that technologies are aligned with public sector priorities and competencies.