In Neo-Colonialism, the last stage of Imperialism, Kwame Nkrumah, a pioneering Pan-Africanist and one of the founding fathers of independent Ghana warned that a state can have nominal independence, while its economic and political policy are dictated externally. According to him, this phenomenon called neo-colonialism does not raise living standards or benefit the countries that have been subjected to it. He posits that control over government policy in the neo-colonial State may be secured by payments towards the cost of running the State, and by monetary control over foreign exchange through the imposition of a banking system controlled by the imperial power. This exchange is not exclusive to political systems. The rapid evolution of computer technology has created complex socio-political consequences, some of which resemble the dynamics of neo-colonial exploitation.

As the world quickly embraces Artificial Intelligence (AI) as the future, many countries are positioning themselves to benefit from AI’s many use cases in their economies. This creates a new struggle for sovereignty over the data used to create and train Artificial Intelligence Systems. The anti-colonial struggle has shifted from a struggle for the right to control physical territory to a struggle for the right to control the data that is used to build Artificial Intelligence systems.

Data is used to train AI models to recognize patterns, make predictions, and perform tasks accurately; it is the raw material with which AI models are built. Without sufficient and high-quality data, AI models may fail to generalise well to new or unseen situations, leading to poor performance and unreliable outcomes. Sovereignty in relation to Artificial Intelligence is the ability to train, deploy, and operate AI systems without dependence on foreign infrastructure or foreign permission.

To achieve this kind of sovereignty, African Countries must have Computing Infrastructure, Data Sovereignty, and Technical Capacity. Currently, a lot of this is outsourced to Western monopoly powers such as Facebook, Google, and OpenAI. This creates an AI Rent Trap, where African Countries will keep having to pay “rent” in foreign currency to access the services built on data, some of which collected from Africa. An additional problem is what Abeba Birhane calls “algorithmic colonisation”; the desire of foreign monopolies to unilaterally dominate human behavior as raw material. Algorithmic colonisation sees to it that AI models that Africans use are built around Western values, making them unsuitable for the African context. This leaves developing nations in a state of dependence on foreign AI systems that is reminiscent of colonial rule.

The wholesale importation and adoption of AI tools made in the West by Western developers is harmful because the tools may not be applicable to the local context, and reliance on these products can hinder the development of local products. Foreign companies that extract data from the Global South benefit from data intensive ventures such as targeted advertising, yet the source countries do not receive commensurate returns. African countries may also be used as testing grounds for controversial technologies, such as Cambridge Analytica’s misuse of AI to influence elections in Kenya and Nigeria; disrupting electoral processes. AI systems trained on globally harvested data are increasingly integrated into military, surveillance, and strategic decision-making architectures in powerful states. When data originating in Africa contributes to these systems, African societies participate in geopolitical dynamics over which they exercise no control. These are some of the ways in which the AI revolution reflects the exploitative dynamics of neo-colonialism.

In March 2024, an underwater rockslide in the Ivory Coast resulted in some submarine cables being damaged, and the loss of internet connection in the 13 West African Countries that relied on them. While some Service Providers were able to reroute traffic to other cables, the domestic digital economy was crippled for some time; taking with it portions of the country’s banking system, health records, e-commerce, and transportation systems in Ghana. Corporations such as Microsoft also suffered some downtime in some parts of the world, but recovered at a much faster pace. The data is used to make Ghana Cards, Ghana’s national ID is owned by Margins Group Ltd, a company backed largely by Danish and American investors rather than wholly by the Government of Ghana. This shows the importance of having local alternatives to critical digital infrastructure beyond the hyper-scalers that dominate the local and international market.

There have been various attempts to create strategies and legal frameworks to govern digital technology in the past decade, the most recent being a slew of policies and strategies to govern the deployment and use of Artificial Intelligence use. It appears that these policies are modelled after European regulatory frameworks such as the European Union (EU) AI Act, the principles in the EU General Data Protection Regulation (GDPR) and the language of trustworthy AI. The problem with this approach is that solutions devised in one culture may not transfer well to another. For instance, an eHealth system designed and developed in Europe is more likely to suggest mammograms as an effective means of detecting breast cancer early whereas in Sub Saharan Africa, self-examinations and clinical breast examinations have been proven to serve better. Legal and policy frameworks must reflect the values, norms, and interests of the societies that rely on them.

As African Governments and regional bodies struggle largely with Western Monopolies for the right to control the digital economy, this article is not advocating an isolationist form of data sovereignty that vests the State with sweeping power over the digital economy to the detriment of its people. Digital Sovereignty should mean that African people have the infrastructure to train Artificial Intelligence models to solve unique local problems, and that where our data is used to train foreign models, we do not fall deeper into the rent trap.

Rwanda has been the most consistent in ensuring that it has computing infrastructure, data sovereignty, and technical capacity. It currently has 3 data centres and mandates all government institutions to host all government data in its National Data Centre. Rwanda’s approach isn’t about AI specifically; it’s about digital infrastructure sovereignty as foundation. AI sovereignty, if Rwanda achieves it, will be built on that base. Other countries such as South Africa and Nigeria have technical capacity and market size, however they lack the political will to devote resources into developing sovereign digital infrastructure. In 2017, Estonia operationalised the world’s first data embassy. It was based on an agreement that allowed Luxembourg to host Estonian data and information systems on a Luxembourg government operated data centre space. This arrangement ensured that Estonia could mirror and protect data abroad in a manner consistent with the legal protections that diplomatic missions have. Though establishing these data embassies may involve complex legal and diplomatic foundations, the African Union and other regional bodies may help create a level playing field for member states and remove some of the legal barriers involved. The process may be enhanced by existing treaties like the Africa Continental Free Trade Area (AfCFTA) Agreement. Data embassies may also support the African Union’s Data Policy Framework, which envisions preserving data ownership and building interoperable systems to foster a secure, integrated continental economy. Using regional frameworks to build interoperability reduces or almost eliminates the AI “rent” that countries and users pay; and increases the bargaining power that countries and users have against Western monopolies.

On the 69th anniversary of Ghana’s Independence from British Colonial rule, it is important to assess the steps that the country has taken to fight for Digital Sovereignty. 2026 has seen the government renew its efforts to build a sovereign AI future and move beyond being a digital dependency zone where foreign entities control the data stack. Legislation alone will not guarantee sovereignty. The true test is procurement. When a Ministry signs a cloud contract, where does the data reside? Under which jurisdiction? Who has ultimate technical control? Can the system function if diplomatic relations sour or sanctions are imposed? Digital sovereignty is not declared in Parliament; it is negotiated in infrastructure contracts.

We must move beyond Nominal Independence where we fly our own flags but run our societies on borrowed data and rented servers. To be sovereign in the age of Artificial Intelligence is to own the data of our people and the infrastructure of our State. By building local capacity, embracing regional data embassies, and standing behind the collective shield of the African Union, we can ensure that our digital evolution is not a new form of annexation, but the ultimate fulfilment of the Pan-African dream.

1957 gave us the Flag; 2026 must give us the Server. On this Independence Day, let us continue to resist oppressors’ rule with all our will and might forevermore!

Key Takeaways:

  • Independence is an ongoing historical process, and a modern state cannot legitimately claim genuine sovereignty if it remains a passive digital colony..
  • Governments must cease writing regulations for technology they do not own and instead invest heavily in domestic compute clusters and localised infrastructure.
  • Combating algorithmic colonialism requires aggressively ensuring that AI tools are built by and for local communities to prioritise localised socio-economic needs.
  • Public sector leaders must demand equitable international partnerships that guarantee comprehensive technology transfer and absolute local data ownership before signing contracts
  • Navigating the future requires embracing true sovereignty by owning the foundational digital architecture rather than relying on isolationist policies that restrict beneficial global engagement