On 26 May 2026, I was invited to speak at the Graduate School of Business Leadership (SBL) at the University of South Africa (UNISA). I titled the presentation “Rooted to Rise: African Knowledge as the Catalyst for Digital Transformation.” I began with a simple question: think about the theories you used in the last research paper you wrote. How many of them were developed in Africa, about African societies, by African scholars? It is a question I increasingly find difficult to ignore, not because theories developed elsewhere have no value, but because Africa remains too often a place where theories are applied rather than a place from which theory is expected to emerge.
This matters particularly in research on technology. Information systems, digital transformation, artificial intelligence and platform-economy research are filled with concepts that have travelled remarkably well. We study technology acceptance, innovation diffusion, organisational change, user behaviour and digital platforms through frameworks developed largely in North American and European institutional settings. These theories can be useful. I use and engage with many of them myself. The difficulty begins when their assumptions become invisible. A particular understanding of the individual, organisation, worker, household or community gradually becomes treated as though it were simply an understanding of human behaviour itself. African reality then appears as the variation requiring explanation.
One of the arguments I made at UNISA was that Africa should be more than a test site for theories generated elsewhere. African experience should also be capable of generating concepts that travel in the opposite direction. This requires taking indigenous African knowledge seriously as a source of theory rather than treating it as cultural background to be added after the supposedly serious theoretical work has already been done.
Ubuntu offers a useful example. The expression commonly associated with Ubuntu, umuntu ngumuntu ngabantu, is often translated as “a person is a person through other persons.” It is usually presented as an ethical idea concerned with community, shared humanity and our obligations towards one another. Those dimensions matter, but its intellectual possibilities extend much further. Drawing particularly on the work of Mogobe Ramose, I discussed Ubuntu at UNISA not simply as an ethical value but as a way of thinking about being and knowing, in which knowledge is relational, contextual and formed through relationships with others.
That has direct implications for how we think about technology. Much of contemporary digital technology is designed around an individual user. There is an account, an individual password, personal data, personal preferences and individual consent. The user evaluates a technology, decides whether it is useful and accepts or rejects it. These arrangements appear technically neutral, but they carry assumptions about what a person is and how that person relates to others.
What happens if technology use is more relational than this model allows? What if trust rests partly on relationships and communities rather than simply on confidence in an institution? What if knowledge is understood as something generated and sustained collectively rather than as the property of isolated individuals? What happens when information collected about one person simultaneously reveals something about a family, lineage or wider community?
These questions become practical very quickly. They matter when we design digital identity systems, health-information systems, artificial intelligence, financial technologies and mechanisms for governing data. Contemporary data regulation, for example, rightly gives considerable attention to the rights of the individual data subject. But some data are inherently relational. Genetic information tells us something about relatives. Digitised indigenous knowledge may affect a community even when a particular individual provides it. Information gathered from farmers, workers or neighbourhoods can generate commercial knowledge whose value exists because thousands of people contributed to it.
Ubuntu therefore encourages us to ask whether individual data rights are sufficient. What might collective data rights look like? Who should have a claim over knowledge created from the aggregated experiences of a community? How should benefit be distributed when collective data produces commercial value? These questions are not relevant only to African societies. They are becoming increasingly important wherever digital systems convert collective human activity into data.
My own research has increasingly led me towards another African concept: Omoluwabi. Rooted in Yoruba moral and social thought, Omoluwabi encompasses ideas around character, integrity, respect, responsibility, hard work and appropriate conduct towards others. It describes something about the kind of person one should strive to become and about how that person exists within a social world.
At first glance, this may appear quite distant from technology. I think it takes us directly into some of the most interesting questions about the digital economy.
Digital platforms now organise relationships between people who may never meet. A freelancer in Lagos can work for a client in London, Toronto or New York. A seller can build a business almost entirely through Instagram or WhatsApp. A worker’s reputation can be compressed into a rating. Communication becomes textual and often highly transactional. Platforms determine how people are evaluated, how disputes are resolved and sometimes even how much they are paid.
Research on platform work has given us important concepts for understanding these relationships, including algorithmic control, surveillance, autonomy, precarity and platform governance. Yet workers do not enter digital markets without histories, values or moral expectations. They bring their social worlds with them.
A worker’s understanding of fair behaviour, what it means to be trustworthy, how one communicates respectfully, what obligations exist towards a client, when maximising income becomes inappropriate, or how reputation should be protected may be informed partly by cultural and social traditions that existed long before the platform. Omoluwabi therefore gives us another way of thinking about digital work. Instead of asking only what the platform does to workers, we can also ask what moral frameworks workers bring into the platform and what happens when those frameworks encounter the economic logic embedded in digital systems.
There can be significant tensions. A worker who understands integrity in terms of honest effort may encounter an hourly payment system whose mechanisms for monitoring work create new ethical dilemmas. A culturally grounded expectation of relational communication can come up against platforms designed around short and impersonal exchanges. Understandings of respect, responsibility and reputation may sit uneasily alongside anonymous ratings and automated dispute procedures.
Technology has not removed culture from economic exchange. It has changed the environment within which culture is expressed, negotiated and sometimes challenged.
This is why I am increasingly dissatisfied with the way we casually say that “context matters.” Of course context matters. But often, when researchers make that statement, the underlying theory remains untouched. We take a framework developed elsewhere, apply it in an African setting, observe behaviour that does not quite fit the model, and then describe “context” as the explanation.
Africa becomes the contextual variable.
There is another possibility. Perhaps the observation is telling us something about the theory itself.
In the UNISA presentation, I described this as a problem of using a map to describe territory that the map was never intended to chart. African innovations can be analysed through frameworks originating elsewhere, African organisations can be evaluated through criteria developed in different institutional settings, and African communities can eventually be described as “not ready” for technologies when the deeper problem may lie in the mismatch between the technology, the theory and the social environment in which it is being used.
The answer is not to reject Western theories. That would reproduce precisely the intellectual closure that needs to be challenged. The argument is for theoretical plurality. Researchers should be able to draw from ideas generated in Europe, North America, Asia, Latin America and Africa depending on what they are trying to understand. Good theories should travel. The problem is that intellectual traffic has historically travelled overwhelmingly in one direction.
African research should be capable not only of confirming, extending or modifying theories generated elsewhere, but also of producing concepts that scholars elsewhere find useful. Ubuntu should not have to remain a theory only about Africa. Neither should Omoluwabi. If Ubuntu helps us think more carefully about collective data rights, relational knowledge or community-centred artificial intelligence, those questions are relevant well beyond the continent. If Omoluwabi helps us understand integrity, responsibility and moral conduct in digitally mediated work, its theoretical contribution need not end at the borders of Yoruba society.
Most theories begin somewhere. Their local origins do not prevent them from becoming globally useful.
Artificial intelligence makes this discussion even more urgent because AI systems do not simply process information. They categorise. They infer. They rank. They predict. Increasingly, they participate in decisions about credit, employment, healthcare, education and access to public services. Every classification requires categories, and every model relies upon some representation of reality.
This raises an important question: whose concepts organise the reality that an AI system is being asked to interpret?
During the UNISA talk, I brought together the work of African and Africa-centred thinkers including Mogobe Ramose, Achille Mbembe, Sabelo Ndlovu-Gatsheni, Abeba Birhane, Oyèrónkẹ́ Oyěwùmí and Tshilidzi Marwala. Their intellectual projects are different, but collectively they force us to confront the assumption that concepts developed within particular historical and cultural circumstances can always be treated as universal.
Oyěwùmí’s work is particularly useful here because it demonstrates how analytical categories imported from one social world can distort another. Her work on Yoruba society challenges scholars to ask whether apparently universal categories actually correspond with the social realities being studied. Artificial intelligence magnifies this problem because categories are no longer confined to books and journal articles. They can be embedded into software and applied to millions of people at extraordinary speed.
African participation in the AI economy therefore cannot mean only increasing access to models developed elsewhere. Africa also needs to participate in defining the languages, datasets, categories, values and assumptions from which intelligent systems are constructed. Otherwise, we risk becoming technologically sophisticated users of systems that understand our societies poorly.
This argument should not be confused with romanticising indigenous knowledge. Everything described as African is not automatically useful, fair or desirable. African intellectual traditions, like every intellectual tradition, must be open to criticism. African societies have their own inequalities, hierarchies, exclusions and contested histories. A concept does not become theoretically valuable merely because it has an indigenous name.
It has to do intellectual work.
It should help explain something that existing concepts cannot adequately explain. Its assumptions need to be made visible. It must be investigated empirically, challenged, refined and, where necessary, rejected. This is precisely why indigenous ideas need to move beyond decorative references at the margins of research. They should be given the same theoretical responsibility that researchers routinely give established frameworks.
At UNISA, I challenged researchers to identify an African concept that genuinely speaks to the phenomenon they are studying, place it at the centre of the theoretical argument where appropriate, subject it to the same rigour they would apply to any established theory, and then publish from it. Indigenous theorising requires intellectual confidence, but it also requires discipline.
There is a larger ambition behind this. African scholarship has spent a considerable amount of time explaining Africa to the world. We should increasingly also produce ideas from Africa with which aspects of the world can be explained.
This distinction is important. Mobile money, platform work, informal digital commerce, community knowledge systems and the improvisational practices of African businesses should not be treated merely as unusual cases waiting to be interpreted through concepts generated elsewhere. They are environments from which concepts can emerge.
During the UNISA presentation, I discussed African digital innovation through examples including mobile money, township entrepreneurship and digital government. The broader point was that African digital transformation should not always be imagined as an imperfect attempt to replicate technological models created elsewhere. African societies are also sites of invention, adaptation and new forms of organising economic and social life.
The same is true of African knowledge systems. Oral traditions, for example, are often contrasted with written knowledge as though they represent an earlier stage of information organisation. Yet oral knowledge systems contain their own mechanisms of redundancy, retrieval, transmission, authority and access. Looking at them seriously may help us ask different questions about how information can be stored, validated, governed and shared.
Ubuntu, Omoluwabi and other African intellectual traditions will not answer every technological question, nor should we expect them to. Their significance is that they enlarge the range of questions we are capable of asking.
The technological future will force societies to think much more deeply about artificial intelligence, identity, privacy, collective rights, work, trust, human agency and the relationship between people and increasingly autonomous systems. These are technological questions, but they are equally questions about society, morality and what it means to be human.
African intellectual traditions have something to contribute to those debates.
For African researchers, this requires a subtle but important shift. Rather than always beginning with the question of which established theory can be brought to Africa, we should sometimes begin somewhere else:
What does African experience allow us to see that existing theory cannot yet adequately explain?
That is where indigenous theorising becomes more than a discussion about representation or decolonisation. It becomes part of the ordinary work of producing knowledge.
And perhaps one of Africa’s most important contributions to the technological future will not simply be using technologies developed elsewhere, or even building more technologies of its own. It may also be changing some of the ideas through which the world understands what technology is doing to human life.
