In mid-May 2026, Pope Leo XIV released his first major social encyclical (a formal teaching letter addressed to the global Catholic Church), Magnifica Humanitas, devoted to safeguarding the human person in the age of artificial intelligence (AI). For those of us who study law, rights, and ethics in the Global South, it is a more useful document than its provenance might suggest – not because it settles the moral questions AI raises, but because it asks them in the right register.
By taking the name Leo, the pope was already invoking Leo XIII, whose 1891 encyclical Rerum Novarum read the upheavals of industrial labour as the “new things” of its age. Leo XIV makes the parallel explicit and turns it on AI. The effect is to move the debate off the ground where much of it has been stuck – between fears of a distant machine apocalypse and the narrow reassurance of technical “bias” fixes – and onto the terrain of power and distribution, here and now. That is a shift African scholars and policymakers should welcome. It reframes AI not as a future technical risk to be managed, but as a present question of who holds power, who is left out, and whose flourishing is foreclosed.
A claim about ownership
Read from Africa, the encyclical’s most striking move concerns ownership. Catholic social teaching has long held that the goods of the earth are “universally destined” for all. Leo extends that principle to data, algorithms, digital platforms, and infrastructure, arguing that when these remain concentrated in a few private hands, a new injustice is produced. For a continent that supplies much of the raw data on which global AI systems are trained – and a growing share of the low-paid human labour that labels and moderates them – while owning almost none of the models, that argument lands with unusual force. To supply the raw material of a system while having no hand in its design is to be made an instrument of another’s becoming, never the author of one’s own.
It also converges, strikingly, with an African ethical tradition. The encyclical grounds human dignity in relationship: a person becomes fully themselves only with and for others. That is close to the principle captured in the isiZulu maxim ‘umuntu ngumuntu ngabantu’ (‘a person is a person through other persons’) – and to a second isiZulu phrase, ‘ngisazophumelela’ (‘I am still becoming’), which holds that personhood is never a finished possession but something achieved, always incompletely, with and through others. The traditions differ in their foundations, but they share an adversary: a logic that measures human beings by their output and treats those who produce little as dispensable. It is what I have elsewhere called ‘flourishing denied’: the systematic obstruction of relational becoming, treated not as private misfortune but as a wrong done. AI, optimised relentlessly for efficiency, is the newest and most powerful vehicle of that logic.
Africa already has the instruments
Where the encyclical should prompt Africa to go further is on implementation. The continent is not starting from nothing. The African Union endorsed a Data Policy Framework in 2022 and a Continental Artificial Intelligence Strategy in 2024, the latter built explicitly around technological sovereignty and a development-first approach. The African Continental Free Trade Area (AfCFTA) Digital Trade Protocol, adopted in 2024, and its annex on cross-border data transfers, adopted in 2025, are the beginnings of a continental rulebook for who may move African data, and on what terms.
The difficulty is that these frameworks remain largely aspirational. Early implementation shows AI investment concentrated in a handful of states, rules at risk of fragmenting into conflicting national regimes, and African languages and faces remaining marginal in the data that trains the dominant systems.
Here the encyclical’s own weakness is instructive. It diagnoses the problem like a structuralist, naming concentrated private power that now outstrips many governments, but prescribes like a moralist, reaching for a change of heart where what is missing is computing power, capital, electricity, and the political will to ratify and enforce. Africa’s frameworks risk the same gap: they proclaim sovereignty over data without yet building the material capacity to exercise it. Moral clarity and continental ambition are both necessary. Neither, on its own, conjures a data centre or compels a multinational.
Useful without being binding
None of this requires agreement with the whole of the document. It is a religious text, and its account of human dignity, of the body, and of the proper “limits” of the human rests on a particular anthropology that secular, feminist, and other readers will not all share. The dialogue it invites has real boundaries. But one need not accept its metaphysics to find its central reframing valuable: that AI is, before it is anything else, a question about power, dignity, and the common good.
That reframing is the encyclical’s gift, and also its challenge to this continent. Africa has begun to write the rules. The harder task is one no encyclical and no communiqué can perform on our behalf. It is to build the capacity, and contest the distribution of power, that would let those rules mean something – and to insist that the measure of any AI order is not efficiency but flourishing: whether it widens or forecloses the relational becoming through which people come fully into themselves. To borrow the document’s own image of a city rebuilt by many hands, the wall is barely above its foundations. We are, in the fullest sense, still becoming – ngisazophumelela – and on this continent that is less a consolation than a summons.