Algorithmic Sovereignty in Practice: A West African Model for Regulating High-Impact AI Systems

  • Emmanuel Nwoye Breatha LTD/GTE, Port Harcourt, Nigeria
Keywords: Algorithmic sovereignty, AI regulation, high impact AI system, AI accountability, West African governance

Abstract

Artificial intelligence is increasingly shaping public and private decision-making across West Africa, particularly in sectors such as Education, healthcare, financial services, recruitment, agriculture, elections, border management, and public administration. However, regulatory responses in the region remain fragmented and largely centred on data protection, innovation policy, and broad ethical commitments. This paper argues that the most urgent AI regulatory gap in West Africa is not simply the absence of AI-specific legislation, but the absence of a regional mechanism to identify, document, and respond to AI harms after deployment. Many AI systems used in the region are imported, privately controlled, technically opaque, and trained outside the linguistic, social, economic, and institutional realities of West Africa. As a result, citizens may be subjected to automated decisions without clear evidence of local suitability testing, human accountability, appeal pathways, or cross-border regulatory oversight. To address this gap, the paper proposes the West African AI Harm Passport, a Pan-African policy model for governing high-impact AI systems deployed in Nigeria and the wider ECOWAS region. The proposed framework would require developers, vendors, and deploying institutions to provide verifiable evidence of system identity, data provenance, sectoral risk classification, contextual bias testing, local language and accessibility assessment, human oversight, complaint mechanisms, and post-deployment incident reporting. Unlike conventional AI ethics frameworks, the Harm Passport shifts regulation from abstract principles to traceable accountability at the point of social impact. The paper concludes that such a model could position West Africa as a global contributor to context-sensitive AI governance by transforming the region from a passive recipient of imported AI systems into an active regulator of algorithmic Power.

Published
2026-07-21