New Study Warns: AI Likely To Put Up To 23% of Musicians’ Revenues At Risk By 2028

In the report on the unprecedented technology, APRA AMCOS details the results of their study, drawing from a survey of 4,274 member songwriters, composers, and publishers, alongside expert interviews and existing revenue-distribution data.

Factoring in the findings from the research, the Sydney-headquartered organization unit (consisting of two copyright management organizations) arrived at the already-mentioned 23% ‘potential damage’ figure.

Looking solely at the music markets in Australia and New Zealand, this percentage, if proven true, could translate to approximately $152.97 million (AU$227 million) in 2028 alone, representing a decline of $349.82 million (AU$519 million) between 2024 and 2028.

At first sight, as the report further details, the revenue risk is not deterring music professionals from exploring generative AI. According to the survey, 38% of participants confirmed using “AI in their work with music and creation in general,” with 5% capitalizing on AI “consistently or nearly always.”

On a more positive note, 27% of respondents stated they have consistently chosen not to use AI for monetization, with 20% of them avoiding the tools altogether. Additionally, only 14% of the surveyed members admitted to using AI directly in their “creative activity with music.”

On the other hand, a significant portion of respondents admitted to using AI in various capacities. A concerning 56% admitted to using AI for generating lyrics, 54% conceded to utilizing the tools for “ideation/to explore new musical horizons,” and 34% and 28% said they use AI for vocal synthesis and sound synthesis, respectively.

Interestingly, opposition to AI remains strong within the music community. Only 8% of respondents held a somewhat positive or very positive perception of AI, while a total of 82% (equalling 3,504 individuals) expressed concerns over the increasing availability and impact of AI, worrying that it “could lead to music creators no longer being able to make a living from their work.”

As AI continues to influence the industry, legal battles and litigations over music copyrights, identity theft, and likeness laws are becoming more frequent. We should not forget about the recent lawsuit filed by the major record companies against AI music start-ups Udio and Suno, alleging that their generative models had been trained on protected materials without permission.

Some expect that a similar situation may arise with SkyMusic 2.0, a new AI music model established by the Chinese tech giant Kunlun, as questions persist about the sources used for training the model.

Though the report’s projections are based on estimates, it touches upon some alarming reality concerning the role generative AI plays in the music industry. Perhaps it can serve as yet another ‘wake-up call’ for music professionals and legal authorities to take further action to protect human creativity and the interests of artists and other creators.