Fieldwork Content
Early observations to guide interviews and workflow capture.
The purpose of these notes is to make the next round of research more specific: what to ask, what to log, which screenshots matter, and where creator agency appears in the process.
Working Note 01
The upload flood changes the research object.
Platform data now makes it hard to treat AI music as a niche creator behavior. Deezer reported nearly 75,000 fully AI-generated tracks uploaded daily in April 2026, or about 44% of daily uploads. Because Deezer also reports AI-generated music as only 1-3% of streams on its service, the cultural question is not just production volume. It is sorting: how listeners, recommendation systems, distributors, and artists distinguish intentional creative work from bulk upload behavior.
Deezer Newsroom, April 2026Working Note 02
Creators describe AI music work as iteration, not one-shot authorship.
The Suno Creators Ultimate Handbook frames the core loop as brief, generate, repair, and finish. That matters because it moves authorship away from a single prompt event and toward repeated human judgment: choosing keepers, identifying which layer failed, repairing stems or sections, testing versions, and documenting the release folder. Fieldwork should therefore capture decisions, not only final audio artifacts.
Local EPUB source: The Suno Creators Ultimate HandbookWorking Note 03
Tool surfaces are converging around audio-to-audio workflows.
Suno documents audio uploads as a way to transform a recorded idea into song material, including longer uploads for paid plans. Udio describes uploaded audio as source material for extension, inpainting, sessions, remixing, and style matching. These are not just text-to-song boxes. They are becoming revision environments where creators bring rough demos, natural sounds, instrument sketches, or existing owned material into an editable generative loop.
Suno Help + Udio HelpWorking Note 04
Rights language is becoming part of the interface.
Udio's audio-upload help text asks users to confirm they own rights to uploaded audio and identifies commercial music or copyrighted tracks as material not to upload. The public Suno/Udio litigation from record companies shows why this interface language matters: legal pressure is shaping how tools describe training, upload permissions, and creator control. Tool analysis should track where rights warnings appear, how visible they are, and whether they change user behavior.
Udio Help + RIAA litigation statementWorking Note 05
Disclosure is now a release workflow, not an ethics footnote.
YouTube requires creators to disclose realistic or meaningful altered/synthetic content, and its examples include synthetically generating music. For AI-native music videos, disclosure becomes part of distribution craft alongside title, description, metadata, thumbnail, credits, and platform rules. The release culture study should record what creators disclose voluntarily, what platforms require, and what audiences can actually see.
YouTube Help