ARTIFICIAL INTELLIGENCE ADOPTION AND PUBLIC SERVICE DELIVERY: A CRITICAL EVALUATION OF GOVERNMENT IMPLEMENTATION, DATA GOVERNANCE, AND INSTITUTIONAL REALITIES IN NIGERIA

The strategic adoption of artificial intelligence (AI) and AI-adjacent algorithmic systems within public sector governance is increasingly framed by international development agencies and national governments as a transformative mechanism for enhancing efficiency, transparency, and administrative responsiveness. In Nigeria, successive post-1999 civilian administrations have aggressively pursued digital transformation through foundational e-government initiatives, biometric identity databases, treasury consolidation, and automated electoral management systems, culminating in recent explicit national AI policy frameworks. However, the operational effectiveness of these high-tech interventions remains deeply contested within an institutional landscape characterized by persistent infrastructural deficits, endemic administrative rent-seeking, fragile legal oversight, and low public trust. This paper presents a rigorous qualitative analysis evaluating the extent to which government implementation of AI and AI-adjacent technologies has modified public service delivery in Nigeria. Grounded in a synthesized theoretical framework combining the Technology–Organisation–Environment (TOE) model, Heeks’s Design–Reality Gap model, and the Public Values Management (PVM) paradigm, this study critically examines the legal, structural, and socio-political determinants of technological adoption. The findings reveal that the primary impediment to public service delivery in Nigeria is not the absence of advanced algorithmic technology, but rather the structural misalignment between normative technology designs and the institutional realities of the state bureaucracy. Without robust data governance, algorithmic accountability, and institutional integrity, AI deployments risk automating existing administrative dysfunctions, entrenching socio-economic exclusion, and facilitating novel forms of tech-mediated rent extraction. The paper concludes with an actionable institutional framework for achieving responsible, transparent, and contextually grounded AI adoption in Sub-Saharan African public administration.