A new investigation has reignited concerns over the use of copyrighted music in artificial intelligence training, with claims that songs from Australian and New Zealand artists have been used without permission to build AI datasets. The findings have intensified calls for stronger copyright protections and licensing rules for creators.
The investigation, highlighted by APRA AMCOS, identified hundreds of songs by leading Australasian artists appearing in datasets used by AI developers. The organisation says the works were included without licences, consent or payment to the songwriters and musicians who created them.
Among the artists reportedly affected are Kylie Minogue, Nick Cave, Sia, INXS and Crowded House. APRA AMCOS says searchable databases now reveal the scale of the material that has been gathered and shared within the AI development community.
The issue extends far beyond Australasia. Recent investigations have found millions of copyrighted songs from artists around the world included in datasets used to train AI music systems. Researchers and industry groups argue that the scale of the data collection raises significant questions about copyright, ownership and fair compensation.
Music industry organisations say creators risk losing substantial income if AI companies are allowed to use copyrighted works without licensing agreements. APRA AMCOS estimates artists and creators across Australia and New Zealand could miss out on hundreds of millions of dollars in future earnings if mandatory licensing frameworks are not introduced.
The debate has particular significance for Māori and Indigenous artists, many of whom already face challenges protecting traditional knowledge, language and cultural expression in the digital age.
Advocates warn that without stronger safeguards, AI systems could absorb and reproduce creative works, cultural material and artistic styles without acknowledging or rewarding the communities from which they originate.
The controversy comes as governments in both Australia and New Zealand consider how copyright law should apply to rapidly advancing AI technologies. AI companies argue that training systems on large amounts of data can fall under existing legal exceptions, while creators maintain that the use of their work requires consent and payment.
The growing dispute is likely to shape future copyright policy across the creative sector as artists, publishers and technology companies battle over who owns the value generated by artificial intelligence.
For many musicians and songwriters, the issue is simple: if AI systems profit from creative works, the people who created those works should share in the benefits.
#AI #ArtificialIntelligence #MusicIndustry #Copyright #APRAAMCOS #Songwriters #Musicians #MāoriArtists #CreativeRights #DigitalRights #AITraining #CopyrightLaw #NewZealandMusic #AustralianMusic #IndigenousCreatives #TechNews #Aotearoa #RadioWaatea #WaateaNews
