An Original Theoretical Framework for Ethical AI Deployment in Nigerian Healthcare Decision Support Systems

AI-based decision support tools in Nigerian healthcare introduce ethical challenges relating to algorithmic accountability, equity across demographic groups, patient autonomy, and the distribution of clinical responsibility that existing biomedical AI ethics frameworks do not adequately address for resource-limited, pluralistic healthcare systems. This study develops an original theoretical framework for ethical AI deployment in Nigerian healthcare decision support. The Nigerian Healthcare AI Ethics Framework (NHAEF) is developed through a five-stage constructive research process: a systematic review of global healthcare AI ethics frameworks; qualitative primary research through 55 interviews with Nigerian clinicians, patients, AI developers, regulators, and ethicists across five geopolitical zones; identification of Nigeria-specific ethical requirements not addressed in existing frameworks (including communal consent norms, traditional healer integration, resource allocation under scarcity, and colonial epistemic justice dimensions); theoretical construction drawing on African ethics scholarship (Ubuntu ethics, communal relational personhood) and biomedical ethics principles; and framework validation through a Delphi consensus process with 18 expert panellists. The NHAEF introduces four theoretical constructs original to the African healthcare AI ethics literature: communal algorithmic accountability, contextual equity calculus, decolonised clinical epistemology, and reciprocal transparency. Each construct is operationalised with practical deployment implications. The framework constitutes an original contribution to global healthcare AI ethics theory, with particular relevance for the global south.

Keywords: AI ethics, healthcare AI, Nigeria, algorithmic accountability, African ethics