Operationalising Knowledge Harvesting through Artificial Intelligence: An African 8-Principle Framework for Recognition of Prior Learning and Skills Development

This paper presents an African Artificial Intelligence (AI) framework grounded in eight foundational principles designed to address the structural misalignment between global AI systems and African knowledge systems. The work forms part of a broader research trajectory focused on bridging the persistent gap between Recognition of Prior Learning (RPL) policy frameworks and their implementation in higher education and skills development. The study advances the argument that African data is not inherently fragmented but is often misinterpreted by global AI models that are unable to process multilingual, oral, and experiential knowledge embedded within communities. It further demonstrates that language, culture, and lived experience are already structured within human systems, though largely excluded from formal digital infrastructures. As a result, significant knowledge—particularly within informal sectors—remains unrecognised and economically underutilised. The framework introduces eight principles that reposition AI from a system of efficiency and automation to one of recognition, inclusion, and empowerment. These principles address the embedded nature of language in system design, the misclassification of African data structures, the prioritisation of human capability, the loss of knowledge through conventional assessment systems, the role of AI in enabling economic participation, the amplification of systemic bias, the gap between policy and implementation, and the necessity of data sovereignty. This work contributes to emerging discourse on AI, epistemic access, and data sovereignty by positioning Africa as an active co-creator of knowledge systems. It provides a conceptual foundation for the development of a Sovereign AI Knowledge System (AI-EKS) that integrates education, industry, and informal economies. Keywords: Artificial Intelligence, Africa, RPL, Data Sovereignty, Multilingual Systems, Knowledge Systems, Epistemic Access