Inside Kampala’s recording studios, AI has moved from a distant idea to a working part of daily music production — helping Ugandan artists write lyrics, craft melodies and polish tracks in a fraction of the time and cost it once took, while opening professional-grade output to producers who could never have afforded it. It has also, in the same stride, forced a harder conversation about what makes music genuinely human and who owns the copyright on the result.
Among the Ugandan producers already integrating AI into their workflow is Benon Mugumbya, who works as a musician, songwriter, producer and filmmaker. He describes feeding a simple melody into an AI system as a prompt and receiving what he calls a full song back within one or two minutes.
But Mugumbya, despite his enthusiasm for the speed and reach of the tools, argues that the outputs still fall short of the human ear’s expectations. AI-generated tracks, he says, are recognisable precisely because they sound too clean — perfect vocals, crystalline production, no rough edges. “It’s still missing that soul,” he says. His preferred mode of working is augmentation rather than replacement: the AI should help an artist extend what they already have, not generate a whole track from scratch. The human touch, in his framing, is the point.
That tension between technological capacity and human feel is now central to a wider debate inside Uganda’s music scene, and it lands against a set of legal questions that copyright regimes across the region are still working through. As AI systems become sophisticated enough to imitate individual voices and complete musical styles, Ugandan artists have joined a global chorus of creators raising the alarm on unauthorized use.
Hamza Ssekamate, a lawyer, argues that artists already have some tools to push back. Creators, he says, hold moral rights over the integrity and use of their content, and economic rights over its commercial exploitation. When either is breached — for example, when an artist’s work is scraped, cloned or resold without permission — the recourse, at least in principle, is the courts. “You have the right to go to court,” he says, pointing to litigation already unfolding in the United States and beginning to surface in Uganda.
But Ssekamate is candid that the law has not kept pace with the technology. AI, he says, “is here to stay,” and Uganda’s legal framework will need to be updated to address the grey areas — particularly around ownership of AI-generated work, licensing of training data, and enforcement against unauthorized voice or style cloning — if artists are to be able to monetize their work on defensible terms.
The Ugandan conversation echoes debates already playing out across the continent. iAfrica has previously covered the South African copyright framework’s struggle to keep up with AI-generated music, and the AI & African Music project run by the Wits Innovation Centre and the MIND Institute has spent the past year building AI tools designed for African languages and tonal systems — including Zazi, a co-creation platform developed by South African artist Umlilo and Ghanaian engineer Gideon Gyimah that handles stem isolation, voice cloning, track merging and AI-assisted mastering with African music in mind. At the Cape Verde Music Expo earlier this year, artists across the continent wrestled with the same twin questions of promise and threat.
Uganda’s own AI infrastructure story is on a parallel track. The Aeonian Project — a planned 10-megawatt supercomputer facility positioned as a potential Ugandan AI factory — is targeting a 2026 rollout, and Cassava Technologies has pledged up to $720 million in adjacent AI infrastructure investment on the continent. Whether that compute capacity will end up serving Ugandan musicians — rather than being routed primarily to enterprise and government workloads — is one of the questions that has yet to be answered.
For now, the Kampala studio conversation is where the tension is most visible. Artists like Mugumbya are choosing to treat the technology as a collaborator that leaves space for human decision-making. Whether Uganda’s legal framework catches up in time to protect the ones who don’t get to make that choice — because their voice has already been cloned — is, on Ssekamate’s account, the more urgent policy question.
