AI in Africa Policy Project - AI4D

The dual potential of AI to promote human development, while introducing new risks and harms into society highlights the importance of designing policies that maximise the benefits of AI while mitigating risks. Policies on data protection, privacy, security that build trust, and improve accountability and fairness, are essential. What is the state of Africa’s readiness to adopt beneficial AI? What does AI readiness entail in theAfrican context? Currently, Africa lags behind other regions in terms of policy and regulatory readiness for AI. Kenya andTunisia are the only two countries on the continent that have AI strategies (Oxford Insights, 2020). Morethan half of all African countries lack privacy and data protection laws. Meanwhile there are only 14signatories and five ratifications to the African Union Convention on Cybersecurity and Personal DataProtection (the Malabo Convention). The ongoing discourse on AI Ethics (fairness, accountability and transparency) is mainly driven by multinational corporations or their foundations and commercial analysts pushing a (self-interested) self-regulatory agenda. But there are also more publicly-oriented multilateral organisations, researchers and activists in the Global North grappling with these complex issues. But the norms and principles for ethical and responsible AI can’t be driven entirely by advanced economies, nor by the private sector, particularly in relation to voluntary compliance and self-regulation. AI readiness dimensions can and should have different configurations and structures of power. This, therefore, requires investigation in the African context.