South Africa’s vibrant music scene, from amapiano to hip hop, is facing a growing challenge as artificial intelligence companies are accused of using local artists’ music to train AI systems without proper permission, credit, or compensation.
Artists including Kabza De Small, Nasty C and DJ Maphorisa are among musicians whose works have reportedly appeared in AI training datasets.
The practice, known as data scraping, involves AI systems collecting large amounts of online content, including music and related information, to develop technology capable of creating new synthetic songs.
A cybersecurity audit by Cybrsec Clinique reportedly found Kabza De Small’s name appearing hundreds of times in documented AI training datasets, followed by DJ Maphorisa, Spirit of Praise, Nasty C and legendary musician Thomas Chauke.
Industry experts warn that the use of copyrighted music to train commercial AI models could threaten artists’ income and cultural ownership, especially as AI-generated songs become more common in the global music market.
Creators demand protection as AI transforms music
The debate was highlighted at the Africa Rising Music Conference, where musicians, technology experts and industry leaders discussed how artificial intelligence is changing the creative industry.
A major focus was finding ways to ensure artists receive compensation when their work contributes to the development of AI systems.
The Creative Weight Attribution (CWA) system, developed by AIxchange and the Fraunhofer Institute for Digital Media Technology, aims to measure how much an artist’s work influences AI-generated content. Supporters believe the technology could help create a fairer system for licensing and royalty payments.
Gallo Records has also raised concerns about protecting South Africa’s musical heritage, saying it is taking steps to prevent unauthorised use of its catalogue for AI training.
The company warned that uncontrolled AI scraping could affect both modern artists and historic recordings from South Africa’s music archives.
Legal uncertainty creates challenges for South African artists
South Africa’s copyright laws are still struggling to keep pace with rapid AI development. The country’s copyright framework was created before the rise of generative AI, leaving questions about how ownership and compensation should be handled.
Legal experts say businesses using AI-generated material without clear ownership agreements could face copyright disputes and reputational risks.
The pending Copyright Amendment Bill is also being closely watched, as rights holders argue that commercial AI companies should obtain licences before using creative works for training purposes.
Meanwhile, organisations such as SAMRO are pushing for stronger regulations to protect musicians and ensure creators are rewarded when their work contributes to AI development.
AI debate expands beyond copyright issues
The conversation around AI in South African music is not only about protecting existing songs but also about maintaining trust in music creation.
Recent controversies involving AI-assisted music production have sparked discussions about transparency and whether audiences should know when artificial intelligence has played a role in creating or improving recordings.
As AI technology continues to develop, South Africa’s music industry faces the challenge of balancing innovation with protecting the creativity, identity and financial rights of artists.
Many industry leaders believe the future of African music will depend on creating rules that allow technology to grow while ensuring musicians remain at the centre of the creative process.
