ARTIFICIAL INTELLIGENCE ADOPTION IN NIGERIAN PUBLIC PROCUREMENT: A TAM–INSTITUTIONAL THEORY PERSPECTIVE

Abstract

Artificial intelligence (AI) holds significant promise for transforming public procurement in developing countries by addressing systemic inefficiencies, reducing corruption risks, and improving transparency. However, empirical evidence on the antecedents of AI adoption among procurement officials in Sub-Saharan Africa remains limited, leaving policymakers without the evidence base needed to guide digital procurement reform. This study investigates the determinants of AI adoption intention among public procurement officials in Nigerian federal Ministries, Departments, and Agencies (MDAs), examining the moderating role of institutional pressure on the relationships between individual cognitive predictors and adoption intent. A quantitative cross-sectional survey was administered to 229 purposively sampled procurement officers, Heads of Procurement Units, internal auditors, and project and programme managers across five federal MDAs. Data were analysed using Partial Least Squares Structural Equation Modelling (PLS-SEM) via SmartPLS 4.0. Results reveal that perceived usefulness was the strongest predictor of AI adoption intention (β = 0.429, p < 0.001), followed by perceived ease of use (β = 0.245, p < 0.001). Perceived ease of use also positively predicted perceived usefulness (β = 0.415, p < 0.001), confirming the internal TAM pathway. Institutional pressure partially moderated the findings: it amplified the usefulness–intention relationship (β = 0.112, p = 0.039) but did not moderate the ease-of-use–intention path (p = 0.562). The study concludes that AI adoption in Nigerian public procurement is driven primarily by individual utility assessments, selectively conditioned by institutional context. It is recommended that procurement reform initiatives prioritise value demonstration over regulatory mandates, invest in participatory system design to reduce ease-of-use barriers, and equip practitioners with structured AI literacy training.