In the age of AI, ownership of African culture should be understood less as private control over cultural production or expressions and more as a struggle over consent, representation, compensation, and governance.
By Samson Jikeme
Eight years ago, while studying for my master’s degree in Los Angeles, my copyright professor introduced us to the question of ownership in the digital age. The case study was Napster, a peer-to-peer music sharing platform that disrupted the global music industry in the early 2000s before being shut down following a series of high-profile lawsuits over copyright infringement. Napster did more than enable piracy. It forced a reckoning. It became the focal point of debates around authorship, ownership, legal reform, and the economic sustainability of music in an era of rapid technological change. For the first time at scale, the question was not simply who created music, but who controlled its distribution, and who ultimately captured its value.
Today, in 2026, we find ourselves in a similar moment but under far more complex conditions. The technology at the centre of this new reckoning is Artificial Intelligence. Unlike Napster, which disrupted the distribution of finished works, artificial intelligence operates at a much deeper level. It does not merely share culture, it learns from it. By training on vast amounts of data scraped from the internet, AI systems can reproduce styles, patterns, and forms of cultural expression without the permission of creators or contributors. This marks a fundamental shift. And the question is no longer, who owns a song, film, or text, but who owns the underlying patterns from which culture itself is produced.
To properly engage the question at the heart of this essay, we must first define what cultural production means.
African culture is not a fixed object, but a living system composed of oral traditions, languages, music, visual art, performance, and increasingly digital expression. Cultural production, therefore, refers to the processes through which these forms are created, circulated, and interpreted. It is a system of meaning, memory, and expression.
Consider the work of an artiste like Wizard Chan, whose music incorporates ancient chants drawn from his community in Rivers State. These chants are not merely aesthetic elements; they are rooted in historical oral traditions and carry cultural, spiritual, and communal significance. If an AI system is prompted to generate similar chants or vocal patterns, it does not access this history. It draws instead from data, learning from existing recordings, including those of artistes like Wizard Chan and reproduces the sonic pattern without its cultural grounding.
What is produced may resemble the form, but it is severed from its origin. What is lost is not only attribution, but historical continuity. The lineage of the chant, the community from which it emerges, and the cultural meaning embedded within it are displaced, even as the pattern itself is reproduced.
In the age of AI, ownership of African culture should be understood less as private control over cultural production or expressions and more as a struggle over consent, representation, compensation, and governance. The shift is now from ownership of works to ownership of patterns.
The central argument of this essay is that Africa is entering this transition without the institutional frameworks required to define and protect these new forms of ownership, even as the technologies driving them continue to evolve at an accelerating pace. This shift is no longer theoretical.
A friend of mine recently published a book on love and released a companion music album alongside it. Having known him for years, I was aware that singing wasn’t his strongest artistic medium. Curious about the process behind the album, I asked how it was produced, and he told me he had used Artificial Intelligence. As I listened, I noticed that the voice on one of the tracks bore a striking resemblance to Enya. Not in imitation of a specific song, but in tone, texture, and emotional register. What was reproduced was not work, but a pattern.
As an entertainment lawyer, I have increasingly encountered artistes and film producers who admit that it is now cheaper to collaborate with AI or train it to generate the kind of vocals or creative elements they need than to engage human collaborators. But these systems did not arrive at this capability independently. They are trained on vast amounts of data scraped from the internet—data derived from existing cultural production.
There are AI tools specifically designed to serve artistes and filmmakers at a lower cost. But apparent efficiency masks a deeper shift. What is being traded is not simply labour, but control. Value is increasingly concentrated in the hands of those who build and own these systems, while the creators whose work underpins them remain unacknowledged and uncompensated. Copyright law was designed to protect works. It was never designed to govern the extraction and reproduction of patterns.
A similar dynamic can be observed in the circulation of a popular Yoruba gospel song, “Emi Awon Woli”. The first time I heard the acoustic version of the performance, what stood out was the depth and texture of the vocal delivery. Seeking to identify the performers, I traced the recording, only to discover that what I had assumed to be a human performance had in fact been generated by artificial intelligence. There was no immediate indication that the voice was synthetic. The performance carried the tonal complexity and emotional register typically associated with trained vocalists rooted in cultural and spiritual traditions.
What is significant here is not simply that AI can imitate performance, but that it can do so convincingly enough to obscure the distinction between human and machine-generated cultural expression. In such instances, authorship becomes difficult to trace, and the relationship between cultural production and its origin begins to weaken. This is where the urgency of the present moment becomes clear.
Many African copyright regimes were built before generative AI and do not clearly regulate AI training datasets, which leaves creators exposed. More fundamentally, it reflects a deeper misalignment. Western copyright frameworks, in isolation, are therefore insufficient to capture the full moral and cultural stakes of ownership.
AI systems do not merely replicate content; they can reproduce and amplify existing inequalities. When African cultural works are absorbed into training datasets without disclosure or compensation, value flows outward while control diminishes locally. Cultural labour becomes invisible at the very moment it is being most widely utilised. At the same time, these systems risk distorting, flattening, or stereotyping African cultures—particularly when they are trained on incomplete or biased data, and when African voices are absent from their design and governance.
What is required, therefore, is not simply incremental reform, but a rethinking of the frameworks through which cultural ownership is understood and enforced. This includes mechanisms for disclosure, licensing, compensation, and provenance tracking. But more importantly, it requires the development of institutional structures capable of defining how African cultural data is used, valued, and governed in the age of AI.
Samson Jikeme is an entertainment lawyer, cultural critic, and Institution builder. He’s also Editor-in-Chief of Afrocritik. Find him on X: @sjikeme and IG: @sam_culture
Cover photo credit: Google Blog