Amid the clamour of nations entering the global race for artificial intelligence, Africa moves out of step, rich in everything the world covets but often absent where the rules are forged. The rare earths extracted from its subsoil feed the chips of supercomputers, its metals flow through the circuitry of data centres, and its territories become the backyard of a digital geopolitics conducted without prior consultation. While rumba rings out in Kinshasa and internal quarrels fracture the national fabric, historic agreements are signed at the White House between Kigali and Washington, allowing a foreign power to exploit Congo’s resources in full legality. Without debate, without transparency, without a Congolese voice at the table.
At the same time, on the technical infrastructure front, another part of sovereignty is faltering. AfrINIC, the regional Internet registry for Africa, is buckling under the weight of litigation, legal attacks and culpable indifference. If this institution were to disappear, the continent would return to the oversight of regional registries in the North, undoing decades of autonomous institution-building. It would be a historic setback, not only in technical management but also in vision. For this is not just about IP numbers. It is about digital dignity.
In one of my recent dissertations at the University of Malta, I argued that any ethical governance of artificial intelligence in Africa must free itself from imported paradigms and be rooted in relational philosophies such as Ubuntu. Far from the ideals of an imagined Wakanda or hastily drafted national strategies on glossy paper, the task is to build a grounded framework, slowly and patiently, that carries an African vision of the world. A vision in which African humanities, our languages, our wisdom and our institutions are operational pillars, not symbolic ornaments. As Souleymane Bachir Diagne stresses, universalising does not mean imposing a global norm, but bringing the humanities into dialogue while respecting their differences.
The stakes are clear. We must not allow ourselves to be dominated by narratives from elsewhere. We must not be content to connect our territories without questioning the architectures that govern them. We must not sink into digital dependence when the very ground we walk on holds the material foundations of global artificial intelligence. It is time for Africa to speak in its own languages, through its own institutions, and standing in the full depth of its history.
The tracks of artificial intelligence are laid in Africa
Africa is often described as a continent lagging behind in the digital revolution, yet few observe where that revolution’s material foundations are anchored. The chips that power supercomputers, the batteries that keep data centres running, and the components that feed tomorrow’s artificial brains rest on an invisible but decisive foundation. That foundation is African. Beneath the Democratic Republic of Congo, Rwanda, Zambia and Mali lie concentrations of critical minerals without which artificial intelligence would lack energy. Coltan, cobalt, lithium and rare earths are extracted in conditions that are often violent, always opaque, and far too rarely subject to the genuine sovereignty of the peoples concerned.
What the great powers call infrastructure, Africa supplies in silence. The routes of the digital world are laid across its land, but without its voices. On 26 July 2025, a peace treaty between the DRC and Rwanda was signed in Washington, presented by Donald Trump as a historic turning point. Behind the Oval Office cameras and diplomatic smiles, the text provides in particular for the disarmament and disengagement of armed groups in eastern Congo, without specifying practical arrangements or responsibilities. Most troubling is the growing suspicion that the agreement was built around an implicit concession. Peace may have been traded for secure access to critical minerals. Tomorrow’s silicon may therefore be negotiated amid the silence of today’s peoples.
Africa can no longer be regarded as a mere logistical reservoir for the digital world. It can no longer be reduced to a geography of supply without governance. It is a thinking continent, a continent of systems and visions. It possesses not only the materials to power global artificial intelligence but also the conceptual resources to help write its ethical, political and social architecture. It has what it takes to participate as an equal in making the universal.
As Souleymane Bachir Diagne reminds us, universalising does not mean imposing a single model. It means broadening the universal through the diversity of human experiences. Drawing on its cosmologies, languages and philosophies of relation, Africa must affirm that it is not destined to remain the backyard of the global cognitive revolution. It is already its material foundation. It can and must now become its conceptual soul.
Yet identifying injustices is no longer enough. Denouncing asymmetries, mapping extraction or lining up statistics is no longer enough. Africa can no longer wait to be given a seat at the table. It must design and set its own table. Moving beyond speeches, repeated promises and narratives of victimhood has become imperative. This moment demands bold public policies that tie mining to clear strategies for digital sovereignty. It demands strong regional alliances capable of influencing global AI governance. Finally, it calls for the unequivocal defence of our African technical institutions, such as AfriNIC, which remain the quiet pillars of our digital autonomy.
I argue that Africa should not merely participate in ethical debates on artificial intelligence. It should propose their foundations. Drawing on relational philosophies such as Ubuntu, it can offer AI governance based not only on performance, but on mutual recognition. Governance that prioritises repair over prediction, and interdependence over accumulation. This vision is not a distant dream. It is operational. It calls for rewriting conceptual frameworks, de-Westernising norms, and clearly affirming that real sovereignty begins in ideas before taking shape in machines.
AfriNIC must not fall, because dependence does not forgive twice
At the heart of Africa’s digital infrastructure lies a little-known institution whose disappearance would undermine the continent’s sovereignty ambitions. AfriNIC, the regional Internet registry for Africa, allocates the number resources that are to networks what plots are to land. Without this registry, there would be no IP addresses, no reliable identifiers, no stable architecture for online public services, banks, universities or connected hospitals. At a time when artificial intelligence is becoming a strategic pillar, Africa cannot afford to see such a central instrument collapse.
Allowing AfriNIC to fall under the combined blows of private interests, judicial inertia and political blindness would mean returning to an era of digital dependence on outside authority. It would mean once again delegating management of our critical resources to the Internet registries of the North, as in the days when Africa did not yet have its own voice in the network’s architecture. It would be a historic regression. It would erase 20 years of community, technical, diplomatic and political efforts.
This is not about defending an institution out of a conservative reflex. It is about understanding that certain infrastructures belong to the common good because they organise access, regulation and fairness in the digital environment. AfriNIC is one of them. Its fall would cause a profound fragmentation of Africa’s digital space, a return to dependence and a loss of control over flows, uses and innovations. This would directly affect states’ ability to implement coherent digital public policies. It would jeopardise the ambitions for cognitive independence that we now articulate in the field of AI.
This must be stated firmly once again. Sovereignty cannot be built with institutions under outside control. African data cannot be governed from distant servers, nor digital identities through imported norms. Sovereignty begins with control of registries, norms and protocols. It is expressed in the ability to create, protect and develop technical commons. The Guidelines for Ethical, Inclusive and Responsible Artificial Intelligence in Francophone Africa follow this logic. This document, the product of a collective process, recognises that AI is not only a matter of algorithms, but also of infrastructure, context and justice.
Full translation of the original article. This English edition was published on 12 November 2026. Dates and assessments in the text reflect its original publication. Read the original article