When Thompson Gyedu Kwarkye published his study “We know what we are doing: the politics and trends in artificial intelligence policies in Africa” in the Canadian Journal of African Studies, he entered a debate that rarely reaches public conversation. The research examines how Rwanda and Ghana are building artificial intelligence governance systems that reflect their political histories, economic ambitions and social priorities.
The findings arrive at a moment when Silicon Valley, Beijing and Brussels dominate discussions about AI. Yet Thompson’s work redirects attention to Africa, where governments are quietly negotiating the future of AI against a backdrop of global power, donor influence and contested digital sovereignty. The study also challenges the assumption that African countries are passive adopters of AI governance. Instead, it reveals deliberate political choices that shape how AI regulation and innovation unfold across the continent.
At the centre of this debate sits a provocative question that framed Thompson’s research. Do African countries truly understand their role in AI governance? His answer, based on qualitative evidence from policymakers and stakeholders, suggests that they do. But he also illustrates that knowing what to do does not guarantee that countries can escape the gravitational pull of global technology powers.
Why AI governance has become urgent
Across Africa, the rapid spread of AI tools into sectors such as healthcare, agriculture, finance, and urban management has created a sense of urgency. Governments recognise that AI can improve medical diagnostics, optimise crop yields, increase financial inclusion and streamline public administration. Policymakers also see AI as a catalyst for economic competitiveness and technological leapfrogging.
Yet Thompson’s study shows that serious concerns match enthusiasm. AI models require considerable electricity and stable digital networks. These infrastructural constraints are acute in many countries on the continent where energy systems remain fragile. The study also highlights ethical risks, including algorithmic bias, privacy breaches, and potential job losses as automated tools replace human labour.
This complex balance of opportunity and vulnerability explains why many African governments now prioritise AI governance. The desire to harness AI for development is closely tied to the need to mitigate structural inequalities, protect citizens from technological harm, and avoid repeating the history of external dependency.
The global pressure behind African AI policies
Thompson’s research reveals that the push for AI regulation in Africa is not solely domestically driven. International donors often link funding to ethical technology guidelines. Organisations in Europe and North America promote governance models that emphasise Responsible AI principles, privacy safeguards and transparency obligations.
African countries have also found themselves at the receiving end of rules written elsewhere. For example, policies introduced by global social media platforms and digital corporations have forced states to engage with regulatory frameworks they did not design or contribute to in any way. This experience has heightened the ambition to shape local AI policies rather than adopt external ones by default.
The study identifies the influence of the European Union (EU) in particular. Through what some legal scholars refer to as the “Brussels Effect”, EU rules on data protection and AI ethics are increasingly shaping global norms. Ghana’s emerging AI policy aligns with EU standards in several areas, especially on privacy, transparency and fairness. Rwanda, too, has incorporated strict protections through its Data Protection and Privacy Law, passed in 2021.
Inside the policymaking rooms
Thompson’s research shows that AI policy development in Rwanda and Ghana is not a linear technical exercise. It is a political process involving state agencies, foreign governments, private companies, civil society organisations and global think tanks. Stakeholder engagement workshops, consultations with start-ups, SWOT analyses, and roundtable dialogues are central to how both countries gather evidence and set priorities.
In Rwanda, the process began with workshops led by the Ministry of Information, Communication Technology, and Innovation. Participants included the Rwanda Utilities Regulatory Authority, the Rwanda Space Agency, Smart Africa and Germany’s FAIR Forward initiative. These engagements shaped the seven-pillar framework that underpins the country’s National AI Policy.
Ghana followed a similar path through the Ministry of Communication and Digitalisation. Thompson notes that at least fifty stakeholders were consulted, including researchers, telecom operators, technology companies, and civil society actors. This early phase helped identify strengths and gaps in Ghana’s AI ecosystem and laid the foundation for a 10-year strategy.
AI policies in Africa do not occur in a vacuum but entail formal and informal consultations with political actors, interest groups, and international bodies.
-Thompson Gyedu Kwarkye
Rwanda’s pursuit of capacity and data sovereignty
One of the most striking insights in the study is Rwanda’s emphasis on local capacity building. The National AI Policy mentions capacity more than a dozen times, signalling a deliberate effort to cultivate domestic skills, research and innovation ecosystems. Rwanda aims to reskill young professionals, expand its technical infrastructure, and invest in research centres to advance AI capabilities.
Thompson links this approach to the country’s historical experiences. The legacy of the 1994 genocide and earlier colonial manipulation of identity records continues to shape national concerns around data security and trust. For policymakers in Kigali, data sovereignty is essential for stability and cohesion. This is reflected in the stringent protections enshrined in Rwanda’s constitution and privacy legislation.
The study also shows that Rwanda’s developmental agenda drives its AI ambitions. AI is seen as a solution to structural challenges such as limited healthcare access, agricultural inefficiencies and rural poverty. Policymakers expect AI tools to help transform public services, support economic diversification and strengthen national security through reliable data systems.
Ghana’s broad vision for innovation and investment
Ghana’s strategy takes a different route. Instead of centring local capacity and data sovereignty, the policy adopts a wide-ranging approach that targets digital innovation, economic competitiveness and social transformation. It outlines eight strategic pillars covering machine learning research, public-sector adoption, future workforce skills, and the mitigation of algorithmic harms.
This approach reflects Ghana’s political and economic landscape. The country has maintained a stable democratic system and has invested heavily in digital technologies over the past two decades. These investments paved the way for foreign interest, including the establishment of Google’s first African AI research centre in Accra in 2019. Such developments have positioned Ghana as a potential regional leader in AI research and commercial innovation.
Thompson notes that Ghana’s reliance on the 2012 Data Protection Act (Act 843) demonstrates ongoing efforts to integrate ethical safeguards. However, the policy calls for further updates to align with global practices and strengthen mechanisms for secure data storage. The country’s emphasis on inclusive growth and technology-driven development reflects broad national goals rather than a narrow security-centred vision.
The shadow of Big Tech and external power
Towards the conclusion of the study, Thompson raises a critical concern. While Rwanda and Ghana are taking significant steps in AI governance, the global AI landscape remains dominated by a small cluster of corporations and powerful states. Companies such as Google, Meta, Amazon, Baidu and Tencent possess resources and influence far beyond that of African governments. Their platforms shape data flows, algorithmic standards and market incentives.
This imbalance raises the risk of techno-colonialism, where African countries rely on tools and standards they cannot fully control. Even with strong policies, enforcement is challenging when regulatory power is asymmetrical. Thompson argues that African governments must actively participate in global regulatory dialogues to assert their priorities and protect local interests.
African AI governance, therefore, sits within a geopolitical contest that extends far beyond the continent. Rwanda and Ghana are not passive participants. They are shaping their own frameworks. However, the global tug of war between regulatory regions and corporate power will determine how much autonomy African countries can maintain as AI technologies advance.
Reference
Kwarkye, T. G. (2025). “We know what we are doing”: the politics and trends in artificial intelligence policies in Africa. Canadian Journal of African Studies, 59(3), 437 to 455. https://doi.org/10.1080/00083968.2025.2456619
