A new study carried out for collective management organizations (CMOs) in Europe predicts that the global market for generative AI in music will exceed USD $3 billion by 2028 – a tenfold increase from 2023.
The study also found a great deal of concern among composers and other music rights holders about what AI will mean for artists’ incomes – and nearly unanimous support for payments to rights holders when their music is used to train AI.
Carried out by Goldmedia, a consultancy and research group focused on the media and entertainment industries, the “AI and music” study predicts a 27% potential shortfall in music creators’ revenues by 2028, if no system of remuneration for human-created inputs is put into place.
In France and Germany alone, that would represent a gap of EUR €950 million (USD $1.03 billion at current exchange rates) in music revenue in 2028, and a cumulative total gap of €2.7 billion ($2.93 billion) between today and 2028.
“Traditional revenue streams are under severe pressure from generative AI for many music creators,” the study’s authors wrote.
“An increasingly predatory competition is to be expected for creators, especially in areas where AI is particularly likely to replace existing human-made music.”
The report was commissioned by the CMOs Sacem – France’s Society of Authors, Composers and Publishers of Music, which represents more than 210,000 authors, composers and publishers – and GEMA, Germany’s Society of Authors, Composers and Publishers of Music, which represents the copyrights of more than 90,000 members (composers, lyricists and music publishers) and more than 2 million copyright owners worldwide.
The study, which Goldmedia called “the world’s first” to look at “the impact [that] generative artificial intelligence is having on the music and creative sector,” involved a market analysis, interviews with experts and a survey of 15,073 people who work as authors or music publishers – members of SACEM and GEMA – in France and Germany.
“An increasingly predatory competition is to be expected for creators, especially in areas where AI is particularly likely to replace existing human-made music.”
Goldmedia AI and music study
The survey found that 35% of respondents have already used AI of some kind in their music creation work. Among those under age 35, that number rises to 51%. Among all respondents, 19% were “AI refusers,” while another 26% would “rather not” use AI.
The use of AI in music varies widely by genre and category, the study found, with it being most common in electronic music, where 54% of creators have already used it, in hip-hop (53%), advertising music (52%), background and stock music (47%) and the audiovisual industry (46%).
It’s relatively less common in pop music (40%), rock and metal (38%), symphonic/contemporary and electroacoustic music (36%), jazz/blues/improv (33%), and traditional and world music (30%).
“Artificial intelligence is already present in the daily lives of creators, and in Sacem’s activities,” Cécile Rap-Veber, CEO of Sacem, said in a statement.
The survey found that the relatively high uptake of AI technology among creators is dwarfed by concerns about AI, with 71% of respondents saying they fear that use of AI could make it impossible for music creators to make a living off their work.
There are currently numerous lawsuits against AI developers over their unauthorized use of copyrighted music in training AI models, but no legal framework to guide how courts should resolve those cases.
The European Union is in the final stages of hammering out an AI Act, and it’s unclear whether the final law will prevent AI developers from using copyrighted works without authorization.
“Authors need transparency and control over the use of their works. But most importantly, they need a fair share of the revenues.”
Dr. Tobias Holzmüller, GEMA
The survey found that 95% of respondents want AI developers to disclose any use of copyrighted works in training AI – something that will likely be a part of EU rules – while 90% want AI developers to ask for permission before using copyrighted works in training, and the same percentage believe copyright holders should benefit financially when their works are used to train AI.
Additionally, 89% want AI-generated music to be identified as such – something that platforms like TikTok and YouTube are moving towards.
“We are actively working with industry players to put in place technical solutions to identify AI-generated content. So it’s not a question of slowing down your activity, but of making it virtuous. The figures in the study show that the estimated damage could be considerable for creators,” Rap-Veber said.
“As part of the finalization of the European regulation on AI, we call on the French and German governments not to oppose the implementation of effective transparency requirements for generative AI companies.”
The study forecasts that the global market for generative AI music will hit $3.1 billion in 2028, up from $300 million (or 8% of the total AI market) in 2023, on the back of massive investment into AI, which the study said amounted to $50 billion in Europe alone since ChatGPT came on the scene in November 2022.
“This means that in only a few years the market will reach a size that corresponds to 28% of global music copyright collections in 2022,” the study’s authors noted.
“The rapid development of this technology offers enormous economic potential. This is confirmed by the results of the Goldmedia study,” said Dr. Tobias Holzmüller, CEO of GEMA.
“The works created by authors are the basis for this revolution. Nevertheless, from the perspective of many music creators, the risks so far outweigh the opportunities. This can only change if we now shape the overall conditions in such a way that everyone involved can participate appropriately in the success of this development. Authors need transparency and control over the use of their works. But most importantly, they need a fair share of the revenues.”Music Business Worldwide
