A commissioned forecast puts a price tag on AI music risk

The system

CISAC, the International Confederation of Societies of Authors and Composers, represents collecting societies for more than five million creators worldwide. On 2 December 2024 it published a news release announcing a commissioned economic study on generative AI's effect on music and audiovisual creators. The study was authored by PMP Strategy, a management consultancy; CISAC states the full study and executive summary are downloadable. The release frames the study as the first attempt to estimate, at a global level, how much of creators' income could shift toward generative AI providers.

What the documents establish

The release states the study projects the market for AI-generated music and audiovisual content growing from about €3 billion currently to €64 billion by 2028, with Gen AI providers' own annual revenue in music rising from €0.1 billion in 2023 to €4 billion by 2028. Against that growth, the study models music creators losing 24 percent of their revenue and audiovisual creators 21 percent by 2028, a cumulative €22 billion figure (€10 billion in music) it attributes to unlicensed use of creators' works and to Gen AI outputs substituting for human-made recordings. The study's own executive summary, dated November 2024, states its methodology combined qualitative and quantitative research, including interviews and workshops with more than fifty industry professionals across the value chain, converted into market-penetration and revenue-loss estimates, and confirms the 24 percent music figure directly.

Craft and rights

For a composer or rights holder, this reads as an advocacy document commissioned by the body whose members stand to lose income under its own scenario. CISAC's Director General is quoted saying the study was commissioned to show the value copyright works bring to Gen AI companies. That does not make the arithmetic wrong, but the 24 percent figure is a scenario forecast built for a policy argument, not an audited market outcome. This is an editorial read: treat the study as what CISAC and PMP Strategy believe could happen under current rules, not as proof of present-day loss.

Outcomes and open questions

The forecast is conditioned on an unchanged regulatory environment; CISAC frames the outcome as dependent on choices legislators are making now. Whether the 24 percent risk figure tracks reality by 2028 depends on open questions: whether training-data transparency rules take effect, whether licensing markets for AI training emerge, and whether enforcement changes the volume of AI content in circulation. No interim 2025 or 2026 figures have been published against which to check the trajectory.

• Does a cited revenue-at-risk figure describe a measured outcome or a modelled scenario, and under what assumptions was it built?
• Who commissioned the study, and what regulatory outcome are they arguing for?
• What would have to change in licensing or transparency rules for the projected transfer of value to be measured directly rather than estimated?

The CISAC-PMP Strategy study is best read as a rights organisation's opening argument in a live policy fight, built on a named methodology but not yet tested against actual 2028 outcomes; its value lies in making the scale of the claimed problem explicit rather than in the precision of any single percentage.