Bridging the Legal-technical Gap: AI Surveillance and Data Sovereignty in Africa
TL;DR
The article shows how procurement-led deployment and security exceptions hollow out constitutional rights that the legal texts appear to guarantee, and proposes a hybrid governance model that closes this gap by anchoring transparency and accountability in constitutional baselines.
Abstract
South African and Nigerian law guarantees rights to privacy and due process, but national security carve-outs and proprietary artificial intelligence (AI) systems shielded from regulatory scrutiny can make those protections difficult to enforce. As African governments progressively rely on foreign vendors for AI-enabled surveillance technologies, a central question is how constitutional protections can be meaningfully enforced when the systems governing surveillance remain technically and legally inaccessible to domestic oversight. The theoretical contribution is the concept of the algorithmic sovereignty gap, which names the persistent disjunction between constitutional guarantees and the operational realities of vendor-controlled tools. Drawing on qualitative content analysis of constitutional law, statutes, case law, and investigative reporting in both countries, and extending Pasquale's black-box problem and Couldry and Mejias's data colonialism framework through a postcolonial lens, the article shows how procurement-led deployment and security exceptions hollow out constitutional rights that the legal texts appear to guarantee. The prescriptive contribution is a hybrid governance model that closes this gap by anchoring transparency and accountability in constitutional baselines, supported by mandatory AI impact assessments, audit rights, source-code escrow, data localisation, and procurement transparency.