AI music creation is already part of many creators’ workflows, but adoption varies sharply by audience, geography, and definition of “use.” Surveys from APRA AMCOS, GEMA, SACEM, SGAE, and IFPI show a mix of experimentation, practical production use, cautious optimism, and concern about income. Forecasts through 2028 describe a rapidly expanding market while estimating material risks for human creators.
Contents
- How widely music creators are adopting AI
- How often creators use AI
- Where AI enters music-making
- What creators think AI will change
- What music fans expect from AI music
- Forecasts for adoption and economic impact
How widely music creators are adopting AI
Adoption depends on whether a study asks about AI in any part of creative work or about music-specific activity. The distinction matters when comparing percentages.
In May–June 2024, 38% of surveyed APRA AMCOS music creators in Australia, New Zealand, and overseas said they had used AI in their work with music or creation in general. The same research found that 33% of surveyed APRA AMCOS members used AI technologies in that broad area on a regular basis in 2024. A narrower question found that 14% had already used AI specifically in creative music activity. These figures describe different question bases and should not be treated as interchangeable. APRA AMCOS, “Largest report on AI in music reveals potentially devastating impact for music creators”; APRA AMCOS/Goldmedia, “AI and Music”.
Other rights-organization surveys show similar evidence of meaningful but incomplete adoption:
| Survey population | Period | Reported AI use |
|---|---|---|
| APRA AMCOS music creators, broad music or creation use | May–June 2024 | 38% |
| APRA AMCOS members, broad use on a regular basis | 2024 | 33% |
| APRA AMCOS members, music-specific creative activity | 2024 | 14% |
| GEMA and SACEM members | 2023 | 35% |
| Spanish music creators | 2025 | 34% |
The GEMA and SACEM study reported that 35% of respondents had already used AI in their work in 2023. Its sample included more than 15,000 creators and publishers, and the reported current AI use in music-related work was also 35%. Among respondents younger than 35, the share who had already used AI rose to 51%. GEMA, “Study: AI and music”; GEMA, “GEMA und SACEM stellen Studie zu KI in der Musikbranche vor”; CISAC, “Sacem and GEMA unveil results of study on the impact of artificial intelligence in music”.
In Spain, 34% of surveyed music creators had used AI tools at some point in 2025. A further 17% said they intended to use AI tools soon. That combination suggests a sizeable group had not yet adopted AI but was considering near-term experimentation. The Spanish result is a separate national survey and should not be added to or averaged with the other studies. SGAE, “Estudio SGAE sobre la IA en la música”.
How often creators use AI
The APRA AMCOS/Goldmedia research gives a more detailed view of frequency among members discussing music and creation in general. In 2024, 5% said they used AI always or almost always, while 28% used it quite often. Another 51% used it rarely, and 15% had only tried it once.
This distribution points to a pattern of broad exposure but lighter routine use. Regular use is visible, yet the largest reported group used AI rarely. The “tried it once” category also matters for interpreting adoption percentages: a creator can count as an adopter without making AI a central part of production.
The continuation responses were more favorable than the frequency distribution might suggest. Among APRA AMCOS members who answered the continuation question, 29% said they would definitely continue working with AI and 42% said they were more likely than not to continue. Together, 71% planned to continue either definitely or as more likely than not. These responses describe intentions among respondents to that question, not a forecast of actual future use. APRA AMCOS/Goldmedia, “AI and Music”.
For independent producers, the practical takeaway is that “adoption” can mean several things: a one-time test, occasional ideation, recurring sound design, or frequent end-to-end generation. The available survey numbers support separating those stages rather than treating them as one uniform behavior.
Where AI enters music-making
Among APRA AMCOS respondents who had used AI in creative music activity, the most common reported role was exploratory rather than fully autonomous production. More than 50% used AI for inspiration to write lyrics and test new musical directions.
Other uses were more specific. In 2024, 34% of these respondents had used AI for co-synthesising vocals. Co-synthesising sounds and co-synthesising instruments were each reported by 28%. A smaller but still significant 23% had used AI to compose complete songs with prompts.
These percentages have a specific denominator: respondents who had used AI in creative music activity. They do not represent all APRA AMCOS members, and they should not be read as the share of all music creators composing complete songs with AI. They also describe reported use in the 2024 study, not a measurement of every music-production tool available today. APRA AMCOS/Goldmedia, “AI and Music”.
The ordering of the results is useful for understanding workflow adoption. Inspiration, lyric development, and testing directions can supplement a producer’s existing process without replacing authorship of the finished track. Vocal, sound, and instrument synthesis move AI closer to the arrangement and production stages. Prompt-based complete-song composition represents a broader form of delegation, but it was less commonly reported than the exploratory uses in this survey.
What creators think AI will change
Creator attitudes combine recognition of creative value with concern about livelihoods. In May–June 2024, 54% of surveyed APRA AMCOS music creators agreed that AI technology can assist the human creative process. At the same time, 65% thought AI-use risks outweighed possible opportunities, and 82% were concerned that AI in music could prevent creators from making a living from their work. APRA AMCOS, “Largest report on AI in music reveals potentially devastating impact for music creators”.
The APRA AMCOS member survey also recorded a divided overall view of AI use in the music sector. In 2024, 4% viewed AI use very positively, 37% saw both positive and negative aspects, and 32% viewed it very negatively. These categories do not describe every possible response, so they should be read as selected indicators of sentiment rather than a complete opinion balance. APRA AMCOS/Goldmedia, “AI and Music”.
The GEMA and SACEM research showed similarly strong demand for safeguards. In 2023, 64% of respondents believed the risks of AI use outweighed its opportunities. A much larger 95% wanted more transparency from companies developing AI tools, while 93% wanted policymakers to give greater importance to AI and copyright challenges. In a related GEMA result, 89% of members surveyed believed AI works should be labeled as AI works. GEMA, “GEMA und SACEM stellen Studie zu KI in der Musikbranche vor”; GEMA, “virtuos 01/2024”.
Spanish creator responses focused on market displacement. In 2025, 36% feared being left out of the market because AI was viewed as a way to reduce costs, and 26% were concerned about the progressive replacement of human activity by machines. APRA AMCOS respondents also connected adoption with platform competition: 83% agreed that the rise of AI-generated work made song visibility and discoverability on streaming platforms increasingly important, while 88% believed human-made music should be promoted on streaming platforms. SGAE, “Estudio SGAE sobre la IA en la música”; APRA AMCOS, “Largest report on AI in music reveals potentially devastating impact for music creators”.
What music fans expect from AI music
Creator concerns are matched by high awareness among listeners. In the IFPI Engaging With Music 2023 survey across 26 countries, 89% of music fans said they were aware of AI. Separately, 79% felt human creativity remained essential to creating music. IFPI, “Engaging With Music 2023”.
Among fans aware of generative AI’s capabilities, 76% believed an artist’s music or vocals should not be used or ingested by AI without permission. Seventy-four percent agreed that AI should not clone or impersonate artists without authorization, and 73% said an AI system should clearly list the music it ingested or used for training.
The survey also found broad support for limits. Seventy percent of music fans agreed there should be restrictions on what AI can do, and 64% said governments should play a role in setting those restrictions. These measures describe fan attitudes in the 26-country 2023 survey; the generative-AI items apply to fans aware of those capabilities. IFPI, “Engaging With Music 2023”.
Forecasts for adoption and economic impact
The available economic figures are forecasts and model-based estimates, not observed 2028 outcomes. The CISAC global economic study projected that 24% of music creators’ revenues could be at risk of loss by 2028. It estimated a cumulative €10 billion loss for music creators over 2024–2028, within a combined music-and-audiovisual estimate of €22 billion. The same study projected the global market for AI-generated music and audiovisual content to grow from about €3 billion to €64 billion by 2028. CISAC, “Global economic study shows human creators’ future at risk from generative AI”.
For music specifically, annual revenue of generative-AI providers was projected at €4 billion by 2028, up from €0.1 billion in 2023. GenAI music was projected to represent about 20% of traditional music-streaming-platform revenues by 2028 and about 60% of music-library revenues.
Regional estimates show why geography matters. The APRA AMCOS/Goldmedia study estimated that 23% of music creators’ revenues in Australia and New Zealand could be at risk from GenAI by 2028. It estimated potential GenAI-related damage of AUD 227 million to creators in those countries in 2028 alone and about AUD 519 million cumulatively over 2024–2028.
The GEMA/SACEM Goldmedia analysis projected a generative-AI music market volume of about US$300 million in 2023 and about US$3.1 billion in 2028, implying about 60% annual growth from 2023. It also projected about €950 million in copyright royalties at risk in Germany and France in 2028 without creator participation in AI revenues.
Spain’s SGAE study estimated that AI could reduce Spanish music copyright income by up to 28% by 2028. It estimated about €100 million in lost Spanish music copyright income in 2028 alone and cumulative losses of €160 million to €180 million over 2025–2028. These are study-specific estimates, so they should not be combined with the CISAC, APRA AMCOS/Goldmedia, or GEMA/SACEM totals. GEMA, “Study: AI and music”; GEMA, “virtuos 01/2024”; SGAE, “Estudio SGAE sobre la IA en la música”.
For music producers, these forecasts frame adoption as both a workflow question and a market-structure question. Current surveys show substantial experimentation and continued interest, while the projections emphasize unresolved issues around attribution, permission, labeling, discoverability, and how revenue is shared when AI becomes part of music creation and distribution.