Musician and composer — tasks, one by one
The unit of analysis is the task, not the job title. Each one below carries its direction, whether the judgement rests on evidence or on platform inference, the reasoning, and what it does not establish.
Every task on this page#
Production, library and advertising music
Automating≈ Platform inferenceWriting background and stock music for videos, adverts and media libraries.
This is where generated music competes most directly. In a survey of about 15,000 German and French creators, 52% of those who make advertising music and 48% of those who make library music had used AI, and the study's interviews concluded that a substitution of human-made background music is already taking place. A study commissioned by creators' societies projects AI music at around 60% of music libraries' revenues by 2028.
Studies commissioned by collecting societies that campaign on AI; the substitution finding comes from interviews, and the 60% is a projection.
Writing and releasing songs
Being augmented✓ Evidence-backedWriting, producing and releasing songs to streaming services.
Generated songs are flooding in but are not what people listen to. Deezer says fully AI-generated tracks exceeded 50% of new uploads at their June 2026 peak, about 90,000 a day, but account for only 1–3% of streams, much of that fraudulent. In a controlled listening study with real Suno songs, listeners could not tell AI from human songs in random pairs, though musicians did better. For songwriters themselves AI so far works mostly as a tool: 35% of about 15,000 surveyed German and French creators had used it.
One streaming service's own detector, which it also sells, a study with a mostly Brazilian sample and a societies' survey; none measures songwriters' income.
Scoring film, television and games
Still human-led✓ Evidence-backedComposing and recording original scores for picture.
The US film and television musicians' agreement says the term musicians includes only humans, requires producers to give notice when a score is created using generative AI, and pays the musicians whose recordings were used as prompts. US projections say composers will be needed to write film scores and music for television and commercials.
One US collective agreement, which pays for AI use rather than forbidding it, and a projection; neither measures scoring work.
Performing live
Still human-led≈ Platform inferencePlaying and singing for audiences in venues, events and services.
US projections expect musicians' and singers' employment to show little or no change from 2025 to 2035, with interest in live music supporting demand, and name declining classical attendance and church participation — not technology — as the drags.
A projection that names no AI effect either way; it does not measure live-music income.
Rights, licensing and royalties
Still human-led✓ Evidence-backedLicensing music and collecting royalties, and controlling how a voice or style is used.
Platforms and labels are building rules around generated music. Spotify allows vocal impersonation only when the impersonated artist has authorised it; Deezer demonetises fraudulent AI streams. Warner and Suno settled with an agreement giving artists and songwriters full control over the use of their names, voices and compositions in AI-generated music, and Universal and Udio settled with a platform to be trained on licensed music.
Company announcements and platform policies that can change; they do not show how much royalty income creators receive.