Artificial intelligence is transforming governance, healthcare, education, finance, and public administration across Africa while simultaneously introducing new ethical challenges related to surveillance, digital inequality, algorithmic bias, and data sovereignty. Existing AI ethics frameworks are largely grounded in universal principles of transparency, fairness, and accountability but often pay insufficient attention to the relational, historical, and socio-political realities shaping technological development in African societies. This article addresses this gap by integrating Ubuntu, African liberation theology, and contemporary AI ethics within a phenomenological framework. Drawing on Interpretative Phenomenological Analysis (IPA), it examines how African theologians, AI scholars, and civil society actors understand the ethical implications of algorithmic power and its consequences for human dignity, social justice, and community life. The analysis demonstrates that AI governance in Africa cannot be reduced to technical regulation or compliance with global standards. Instead, participants consistently interpret artificial intelligence through relational understandings of personhood, solidarity, moral responsibility, and collective flourishing. Building on UNESCO's ethical principles while incorporating Innovationology, Noesology, Mode 4 Knowledge Production, and Integral Ecology of Innovation, the article develops a Relational AI Governance framework that situates technological innovation within broader commitments to human emancipation, epistemic pluralism, ecological stewardship, and inclusive development. It argues that African theological traditions are not merely local ethical resources but offer globally relevant insights for reimagining AI governance in ways that promote justice, sustainability, and human dignity.