As artificial intelligence technologies rapidly advance globally, African governments are beginning to develop national AI strategies and policies. However, they face crucial decisions about how to govern artificial intelligence technologies.
A recent analysis of publicly available AI strategies from Egypt, Mauritius, and Rwanda reveals key trends in how African governments are approaching AI policy. How African governments are assembling AI policy notes the following issues.
5 Issues with African AI Policies
The current approach risks creating policies that primarily serve the interests of large tech companies and political elites rather than addressing contextual needs and risks. The policies lack a comprehensive view on how AI could either empower or disenfranchise African populations.
We do not want AI solutions that exacerbate existing socio-economic divides with policies that have:
- Economic and Innovation Focus: The strategies heavily emphasize economic growth, technological advancement, and research & development. Egypt, for example, highlights workforce development and innovation as crucial components of its national AI strategy. There is less focus on ethical considerations or equitable distribution of AI benefits.
- Limited Institutional Engagement: While government institutions are highlighted, there is vague or limited discussion of roles for civil society, local tech companies, and citizens in shaping AI governance. In contrast, Mauritius emphasizes a multi-sectoral approach, involving stakeholders from research institutions and tech start-ups
- Top-Down Approach: Rwanda’s AI policy focuses heavily on the government’s role in driving AI adoption and governance. This top-down, government-led approach may not incorporate diverse stakeholder perspectives. They rarely address the deeper societal impacts of AI, such as distributive equity, or the inclusion of marginalized groups
- Global Discourse Alignment: The strategies often align with global AI discourses and frameworks, prioritizing capitalist and technological agendas, potentially at the expense of locally led and locally relevant priorities.
- Data and Skills Emphasis: There is significant focus on data collection and skills development, but less on data rights or algorithmic accountability. African governments appear to adopt a “wait-and-see” approach.
We Need Locally Led AI Policies
The author notes that African nations appear to be “performing global discourses” rather than developing truly contextualized approaches. Many policies uncritically adopt standardized global frameworks without sufficiently considering local needs, values and perspectives.
As we’ve seen with other ICT4D initiatives, failing to engage local stakeholders can lead to unsustainable or harmful outcomes. To create more inclusive and effective AI policies, African governments should:
- Expand stakeholder engagement to include civil society, local tech startups, and marginalized communities with formal channels for citizen input in AI governance
- Critically examine how AI technologies may impact existing inequalities with concrete enforcement mechanisms to ensure responsible AI deployment.
- Develop contextually relevant ethical frameworks, including incorporating African epistemologies and ethical frameworks.
- Focus on equitable distribution of AI benefits, not just economic growth that explicitly consider how AI systems may concentrate power
- Increase transparency around data collection and algorithmic decision-making with specific responsibilities, accountability mechanisms and oversight structures for all stakeholders in the AI ecosystem.
Market Dynamic of AI Power and Value
One of the study’s most compelling insights is that African governments need to better understand how AI markets actually function – what the author calls “seeing like a market.” This means recognizing how data and algorithmic systems can concentrate power and create new forms of inequality.
The research highlights how AI systems enable unprecedented data collection and classification, often in ways that are opaque to users. This data becomes a tradable commodity, with value extracted by powerful companies while the individuals who generate it may see little benefit.
AI Policy Future View
The research suggests that while current approaches have limitations, there’s still time to develop more effective frameworks that better serve African interests.
Success will require moving beyond simplistic techno-optimism to engage deeply with questions of power, equity and social impact. African governments must resist the tendency to uncritically adopt external frameworks and instead develop contextually appropriate approaches.
African nations can work toward ensuring AI technologies genuinely benefit their societies rather than simply reproducing existing inequalities in new forms.
A synopsis of How African governments are assembling policy in anticipation for data and AI driven techno-futures by Angella Ndaka at the University of Otago
