Research reveals ‘deep’ AI governance gaps in Sub-Saharan African countries

Rwanda stands out for its proactive approach, having established a comprehensive AI strategy and clear data governance laws that prioritize ethical AI use. Other countries, including Ghana and Kenya, show promise but still struggle with the implementation of policies that safeguard local AI development.

As artificial intelligence (AI) becomes a transformative force in global governance, concerns about the fairness and inclusivity of its regulatory frameworks have taken center stage. Historically, AI governance has been shaped by Western-centric paradigms, often marginalizing voices from the Global South.

A recent study, "Decolonizing Global AI Governance: Assessment of the State of Decolonized AI Governance in Sub-Saharan Africa", by Gelan Ayana et al., published in R. Soc. Open Sci.11231994, investigates the extent to which AI governance in Sub-Saharan Africa aligns with decolonial principles. The research evaluates the progress of ten African nations in establishing AI institutions, developing national strategies, ensuring sovereignty, protecting data, and prioritizing local datasets. The findings reveal that while some countries show awareness of decolonization, meaningful progress remains limited, with only Rwanda demonstrating significant responsiveness.

The urgency of decolonized AI governance

AI systems do not operate in isolation - they reflect the political and historical contexts in which they are created. In many cases, AI models trained on Western data carry embedded biases that fail to account for the socio-political realities of Sub-Saharan Africa. The study highlights that without decolonized AI governance, these biases perpetuate inequalities, reinforcing systemic disparities in access to AI technologies and decision-making processes.

The decolonization of AI governance requires a fundamental shift from a Eurocentric regulatory approach to one that centers local knowledge, ethical considerations, and technological sovereignty. The research underscores the need for countries to develop governance structures that prioritize local AI expertise, implement policies that enforce the ethical use of AI, and create data strategies that prevent the unchecked dominance of foreign AI models.

Findings: The state of AI governance in Sub-Saharan Africa

The study assesses ten African nations - Cameroon, Ethiopia, Ghana, Senegal, South Africa, Kenya, Mauritius, Nigeria, Rwanda, and Seychelles - based on five core indicators: the presence of AI governance institutions, existence of national AI strategies, prioritization of sovereignty, implementation of data protection regulations, and the use of local datasets.

Findings indicate that while 80% of these nations are “decolonization-aware,” meaning they recognize the need for decolonized AI governance, only Rwanda is classified as “decolonization-responsive,” actively implementing strategies to counter Western dominance in AI policymaking. Cameroon, on the other hand, is categorized as “decolonization-blind,” showing little recognition of decolonial AI governance principles. Most countries exhibit limited institutional structures, inadequate policies on local data use, and regulatory gaps that leave them vulnerable to foreign AI interests.

Rwanda stands out for its proactive approach, having established a comprehensive AI strategy and clear data governance laws that prioritize ethical AI use. Other countries, including Ghana and Kenya, show promise but still struggle with the implementation of policies that safeguard local AI development. The study calls for increased investment in regional AI partnerships, stronger enforcement of data protection laws, and deeper engagement with local AI researchers to foster sustainable governance models.

Moving forward: Recommendations for decolonizing AI

Decolonizing AI governance is not merely an academic exercise - it is a necessary step in ensuring AI serves all communities equitably. The study outlines actionable recommendations for advancing AI governance in Sub-Saharan Africa. These include fostering regional collaborations to share best practices in AI policy development, encouraging the use of local data to mitigate algorithmic biases, and empowering indigenous AI researchers to take the lead in shaping governance frameworks.

Additionally, governments must prioritize AI literacy programs to build local expertise and challenge the dominance of Western AI frameworks. Investing in AI infrastructure and regulatory mechanisms that reflect African realities will be critical in reshaping global AI governance narratives. The study also advocates for a participatory approach that includes local communities in discussions about AI ethics, ensuring that AI deployment aligns with societal values and cultural contexts.

Ultimately, achieving a decolonized AI governance landscape will require a concerted effort among policymakers, technologists, and civil society actors to dismantle historical inequities and create an AI future that is inclusive, transparent, and just.