In the Zanzibar archipelago of Tanzania, rural farmers utilize an AI-assisted application called Nuru, designed in Swahili, their native language, to identify a devastating cassava disease before it spreads. South African computer scientists have constructed machine learning models to assess the impact of racial segregation in housing. Similarly, in Nairobi, Kenya, AI technology categorizes images from numerous surveillance cameras stationed on lampposts throughout the bustling city center.
The potential economic benefits of AI adoption in Africa are substantial. Estimates indicate that by 2030, Nigeria, Ghana, Kenya, and South Africa alone could accumulate up to $136 billion in economic gains through increased utilization of AI tools.
Presently, the African Union, comprised of 55 member nations, is formulating an ambitious AI policy aimed at charting a distinctly African path for the development and regulation of this emerging technology. However, debates regarding the timing of AI regulation and concerns about impeding innovation could impede progress. Additionally, a deficiency in AI infrastructure may hinder widespread adoption of the technology.
Chinasa T. Okolo, a fellow at the Center for Technology Innovation at Brookings, stresses the importance of establishing regulations to govern AI technologies amidst the continent’s growing AI landscape.
Several African countries have already initiated the development of legal and policy frameworks for AI. Seven nations have crafted national AI policies and strategies, each at varying stages of implementation. Recently, the African Union Development Agency published a policy draft outlining a framework for AI regulation across the continent. This draft includes recommendations for industry-specific codes, standards, certification bodies, regulatory sandboxes, and the establishment of national AI councils to oversee responsible AI deployment.
Although the heads of African governments are anticipated to endorse the continental AI strategy at the AU’s annual summit in February 2025, challenges remain. Countries without existing AI policies or regulations will utilize this framework to develop their own strategies, while others will be encouraged to review and align their policies with the AU’s.
Internationally, significant AI legislation and policies are also taking shape. For instance, the European Union recently passed the AI Act, poised to become the world’s first comprehensive AI law. The United States issued an executive order on AI, while China eyes a comprehensive AI law akin to the EU’s.
Without robust regulatory frameworks, experts warn of potential social harms from AI misuse, including exacerbating inequalities. Moreover, failure to harness AI’s benefits could leave African economies trailing behind.
Some African researchers advocate for prioritizing AI industry development before regulating the technology. They cite challenges such as high infrastructure costs, limited internet access, funding constraints, and a shortage of powerful computers needed for AI model training. Shikoh Gitau, a computer scientist, argues for focusing on innovation and opportunities in Africa before implementing regulations.
Conversely, Chinasa T. Okolo contends that Africa should proactively develop regulations, suggesting reforms to existing laws on data privacy and digital governance to address AI.
African voices have been notably absent from global discussions on AI governance and regulation, emphasizing the need for Africa to shape its own regulatory frameworks. Nyalleng Moorosi emphasizes the necessity of regulation to prevent labor exploitation and safeguard communities against misuse by corporations and governments.
While the draft AU-AI policy doesn’t explicitly address AI use by African governments for national security, it acknowledges potential risks associated with AI. Barbara Glover emphasizes the need for African countries to invest in digital and data infrastructure and collaborate with the private sector to support AI startups and innovation hubs.
Unlike the EU, the AU lacks the authority to enforce policies across member states. Therefore, even if the draft AI strategy gains parliamentary endorsement, African nations must implement the continental strategy through national AI policies and laws.
As machine learning tools continue to proliferate, Moorosi underscores the importance of developing a model for local AI regulation and governance that balances risks and rewards.
while AI holds immense promise for Africa, effective regulation is imperative to maximize benefits and mitigate potential harms.
