| Abstract: |
Artificial intelligence (AI) is reshaping how organisations plan, schedule, forecast and control operations, yet empirical evidence linking AI adoption to both managerial decision quality and measurable operational efficiency remains fragmented. This paper examines how AI adoption intensity influences decision-making quality and operational efficiency, and whether decision quality transmits the efficiency benefits of AI. A cross-sectional dataset of 240 observations from manufacturing, logistics, retail, healthcare and IT/BFSI service firms was analysed using reliability testing, correlation, hierarchical regression, bootstrapped mediation and paired pre/post comparisons of five operational indicators. The dataset is simulated for illustration, with parameters informed by published literature. All constructs showed satisfactory reliability (Cronbach's alpha 0.78–0.86). AI adoption intensity was strongly associated with decision-making quality (β = 0.54, p < 0.001) and operational efficiency (β = 0.44, p < 0.001). Decision quality partially mediated the AI–efficiency relationship, accounting for 39% of the total effect (indirect effect = 0.17, 95% CI 0.09–0.25). After deployment, forecast error fell by 19.8%, order cycle time by 14.4%, inventory holding by 12.0% and unit cost by 6.4%, while on-time-in-full delivery rose by 5.4 percentage points. High adopters outperformed low adopters on efficiency, and data quality and top-management support emerged as key antecedents of adoption. Findings are robust to controls for firm size and sector. The findings suggest that AI creates value mainly by improving decisions rather than by automation alone, and that complementary organisational capabilities are decisive. The paper offers practical guidance for operations managers and a replicable analytical template for future primary studies. |
