CIPESA analysis says Africa's AI governance politics risk leaving states dependent on foreign tech

In an analysis published by CIPESA, author Paul Kimumwe examines why African AI policies and strategies may struggle in practice. He says political will is often lacking, and that this shows up as an absence of clearly defined funding mechanisms. In his account, even progressive frameworks face difficulties in execution because of structural flaws and inadequate money.

Kimumwe observes that, apart from Kenya, most African countries working on an AI strategy or policy set no budget figures for carrying it out. He says Rwanda has a project-level funding framework, while Egypt and Mauritius depend on programme budgets. Kenya's 2025 to 2030 strategy was costed at KSh 152 billion over five years, yet he reports that a review of the 2026/27 national budget found no dedicated allocation, with the Sh8.6 billion for ICT directed mainly at broadband, digital skills and digitising government services.

The author also describes fragmented responsibilities. In Rwanda, he says, duties are divided among a Responsible AI office, the utilities regulator RURA and the cyber security authority NCSA, while in Ethiopia the Ethiopian Artificial Intelligence Institute coordinates and several other bodies hold sector mandates. He warns that this approach carries risks of divergent priorities, power struggles and regulatory fragmentation, and that oversight bodies often lack independence from political appointees.

His recommendations include AI-specific laws, clearly mandated and resourced oversight bodies free from political interference, meaningful stakeholder participation, and a human rights-centred approach, which a 2025 CIPESA study found remains aspirational in many countries.