Suno unveils AI model trained on licensed music amid ongoing copyright lawsuits

Suno unveiled three versions of the model on Wednesday: the base Suno v6 for paying users, described as reliable and steerable for controlled outputs; Suno v6 wild, an experimental model for ideation; and Suno v6 mini, a faster version available to all users. The company plans to retire its older models.

Suno said the new lineup offers greater flexibility, including the ability to edit parts of a song via prompt or lyrics, use text, images or video as references to create tracks, and separate instruments from a sample to create new beats. The startup also plans to introduce a feature allowing users to remix songs, provided the respective artists opt into a new program it is setting up with music labels.

“I think the music ecosystem and our partners are always looking for ways to create more revenue opportunities for their rights holders and artists. So a big part of this release is creating additional revenue streams there,” Jack Brody, Suno’s chief product officer, told TechCrunch.

Suno has settled with Warner Music Group and struck a deal with BMG in recent months, but still faces lawsuits from labels including Sony and Universal Music Group, artists such as Jason Isbell, and users alleging it ignored security while chasing profits. The company said Suno v6 was not trained using the data used to train previous versions of its model.

Last month, Suno said it plans to add watermarks to songs generated on its platform and introduced new download limits based on account tier. Brody noted the company can help manage ecosystem challenges like streaming fraud and mass exportation, but said it is ultimately up to distributors and platforms to govern what content goes on their services.

Why This Matters

Suno's shift to licensed data could set a precedent for how AI music models handle copyright, potentially reshaping the industry's approach to training data.
The outcome of the remaining lawsuits — from Sony, Universal and individual artists — will determine whether AI music generators can operate legally without paying for the vast majority of recorded music.
New features like remixing with artist opt-in may create a new revenue model for rights holders, but the voluntary nature and implementation details will be critical.

Background

Suno is an AI music generation startup whose models allow users to create songs from text prompts. The company has faced a wave of litigation from major record labels and artists who allege it trained earlier models on copyrighted music without permission. Settlements with Warner Music Group and BMG indicated a path toward licensing, but Sony and Universal have not settled. A separate class-action lawsuit from users claims the company prioritised growth over security.

Key Perspectives

Suno and licensed partners (Warner, BMG, Believe): Position the v6 model as a responsible, licensed alternative that opens new revenue streams for rights holders through AI-generated derivative works. Sony Music, Universal Music Group and artists like Jason Isbell: Argue that Suno's earlier models were built on unlicensed copyrighted material, harming artists' rights and compensation. Their lawsuits seek damages and potentially an injunction against Suno's services. Critics/Skeptics: Question whether Suno's licensed data set is truly comprehensive enough to avoid future infringement, and whether the opt-in remix program will attract enough artists to be viable. User concerns about data security also linger.

What to Watch

Whether Sony and Universal's lawsuits proceed to trial or lead to settlements similar to Warner's.
The adoption rate of Suno v6 among paying users and the quality comparison to the older models.
How many artists opt into the planned remix program, and what terms Suno offers rights holders.

ZOTPAPER