Suno and Major Music Labels Clash Over AI Music Sharing — and the Terms of Any Peace Deal

Negotiations between AI music startup Suno and major record labels — including Universal Music Group and Sony Music — over the sharing of AI-generated songs on streaming platforms have hit a significant obstacle, according to a report from the Financial Times. The dispute centers on Suno's interest in allowing users to upload AI-generated tracks to services like Spotify and Apple Music, something the labels are reportedly unwilling to accept as part of any settlement of the copyright litigation that has been ongoing since 2024.

What Suno Wants

Suno's position reflects a business reality: an AI music platform whose outputs cannot be distributed through mainstream music infrastructure is a platform with a severely limited use case. The company has argued that AI-generated music is a legitimate form of creative expression and that users should be able to share it wherever they want. Enabling uploads to Spotify, Apple Music, and YouTube Music would allow Suno to compete with human-produced music for algorithmic promotion, playlist placement, and listener attention — treating AI-generated audio as a first-class creative product rather than a novelty tool.

What the Labels Want

For Universal and Sony, the question of streaming distribution goes to the heart of what any settlement would actually accomplish. The record labels sued Suno (and, separately, Udio) for training on copyrighted recordings without licenses. A settlement that resolves the training-data question but then permits unrestricted distribution of AI-generated music on the same platforms where licensed recordings compete would still damage the core business model the labels are trying to protect. From their perspective, allowing AI-generated tracks onto Spotify is accepting the very competitive threat they went to court to limit.

The Structural Problem

The dispute illustrates why music industry AI negotiations are harder than they appear from the outside. Training licenses and distribution rights are separate questions, but they are not separable commercially. A startup that can legally train on catalog but cannot distribute what it creates has a weaker business. A label that licenses training data but then faces unlimited AI competition on streaming platforms has given up significant leverage for limited gain. The two sides are not just negotiating a dollar settlement — they are negotiating what the AI music industry looks like structurally, and who has power in it.

The resolution, if one comes, will set a precedent that shapes how every other AI music platform operates, what streaming services are willing to host, and how compensation for training data interacts with the economics of music distribution. That is a large set of interests to align, which is why talks are reportedly stuck.