Africa’s Cultural Contributions at Risk in the Age of AI Music
Despite Africa’s rich cultural heritage in world music, its contributions are largely absent from the data used to train generative AI systems. Music industry veteran Charles Goldstuck warns that urgent action is required to secure Africa’s place in the evolving technological and musical landscape. Without proactive measures, the continent could find itself relegated to a consumer role rather than an influential shaper of the future of music technology.
Goldstuck, founder of Goldstate Music and a Wits alumnus, delivered a keynote address at a recent event organized by the Wits Innovation Center and the Wits Mind Institute. This gathering marked the conclusion of the AI and African Music Project, a six-month pilot initiative aimed at pairing African musicians with AI engineers to create customized technological solutions for the continent’s creative community.
During his address, Goldstuck highlighted a troubling disparity within the global music market’s generative AI landscape. He argued that the swift evolution of commercial practices outpaces public policy, suggesting that the industry’s future will be shaped more by commercial negotiations than regulatory frameworks. Notably, French streaming service Deezer disclosed it receives over 60,000 AI-generated tracks daily, underscoring the growing concern over the quality and authenticity of music being produced.
Goldstuck elaborated on the contrasting approaches taken by leading AI music platforms. Suno, which recently secured a valuation of $2.45 billion following a $250 million Series C funding boost, is currently embroiled in copyright lawsuits based on claims of “fair use.” In contrast, Woodio has opted for a more conservative strategy, developing a “walled garden” by signing licensing agreements with major entities like Universal Music Group and Warner Music Group.
Technology Challenges in Attribution and Compensation
While advancements in detection tools have emerged, Goldstuck pointed out that “attribution” technology—capable of tracing AI-generated songs back to specific copyrighted materials—is still in its early stages. He expressed skepticism about whether this technology can be effectively developed and deployed, emphasizing that current compensation models often depend on broad market share rather than accurate usage data.
The scale of challenges facing the music industry is staggering. Spotify has removed more than 75 million “spam” songs from its platform within the past year. Meanwhile, Deezer estimates that up to 85% of streams from AI-generated music are fraudulent, prompting the service to remove these tracks from its royalty pool in an effort to maintain integrity.
During a subsequent Q&A session, the discussion shifted to the concept of “creative sovereignty” for African artists. Goldstuck noted that although Africa has largely escaped the initial repercussions of generative AI—which tends to focus on Anglo-American and Chinese music—there remains a critical lack of technological infrastructure and investment in the region.
“Africa’s cultural contribution to the global music landscape is not reflected in the data that has been compiled thus far,” Goldstuck stated. He called on the creative community to advocate for government involvement in funding technology initiatives related to music.
Professor Christo Doherty, leading endeavors at the Wits Innovation Center, echoed Goldstuck’s call. “AI presents significant opportunities for African musicians, but addressing these challenges requires more than just regulatory oversight,” he affirmed.
Innovative Projects Showcase African Musical Heritage
The showcase event at Wits’ Chris Seabrook Music Hall included live demonstrations of five pilot projects created by teams from seven African nations. The initiative focused on key areas such as music archiving, preservation, and localization of technology tools.
The featured projects included:
- Zazi: This “music digital twin” enables real-time interaction through voice, rhythm, and storytelling, developed by South African artist Umlilo and Ghanaian AI engineer Gideon Gyima.
- Bɛ̀bɛ̀i Engine: Co-created with the Baka community to preserve endangered polyphonic traditions, this performative AI instrument was built by Cameroonian artist Joshua Kroon and Ghanaian AI/ML engineer Emmanuel Apetsi.
- Bina.ai: An AI-driven platform for children’s music and storytelling rooted in African genres, created by Nigerian music strategist Ehinome Ogbeide and Congolese engineer Muhigiri Azuza Albin.
- Heritage in Code: A digital archive and AI fusion engine focused on preserving Africa’s musical instrument heritage, developed by Kenyan deejay Linda Nyabundi and Ethiopian AI researcher Gebregziabihie Nigusie.
- TIMAh AI: Documenting Kikuyu traditional music with a secure archive and community-centered consent governance, this project was led by Kenyan producer Thora Nyamosi and machine intelligence engineer Lawrence Moruye.
As Apetsi remarked during his presentation, “The future of music technology doesn’t have to be imported; it can be developed here, collaboratively, making it distinctively African.” These innovative projects, currently in the prototype phase, are receiving continued technical support from the Mind Institute and Wits Innovation Center.
