AI Generated Music: Xania Monet Sparks Fight for Transparency, Watermarks, and Artist Royalties

AI generated music is climbing the charts, but hidden data, missing labels, and unpaid artists reveal an industry at breaking point.

AI generated music has officially entered the mainstream. Billboard recently confirmed that at least six AI-powered or AI-assisted artists have appeared on its charts across gospel, rock, country and R&B. The real number could be far higher because even Billboard admits it’s “increasingly difficult to tell who or what is powered by AI, and to what extent.”

At the centre of this shift is Xania Monet, an R&B artist created through AI generated music technology. Mississippi-based songwriter Telisha “Nikki” Jones writes the lyrics before using the generative platform Suno to turn them into fully produced tracks. The result – smooth, soulful, and indistinguishable from human performance earned Monet a reported $3 million deal with Hallwood Media, founded by former Interscope executive Neil Jacobson.

The Platform Behind the Voice

The platform powering Monet’s rise is Suno, a leader in AI generated music production. Users input text prompts, and Suno creates entire songs vocals, instrumentation, and mixing included. But one issue remains hidden: no one knows exactly what music trained Suno’s algorithms.

If the system learned from existing copyrighted songs, those works are being reused without permission. Each generated melody might contain echoes of artists who never consented to the process. Until the source of Suno’s training data is disclosed, AI generated music will remain a black box profitable, but opaque.

Why AI Generated Music Must Be Labelled

Transparency is the missing foundation of this new era. Music platforms should be legally required to label every instance of AI generated music or AI-assisted production. Clear disclosure gives audiences the power to choose what they listen to, and it gives human artists the credit they deserve for originality.

Developers of AI composition software should also embed watermarks or provenance metadata into all AI generated music. This would create a permanent trace of how a track was made, enabling streaming services, labels, and collecting societies to identify and classify synthetic works accurately.

Spotify’s AI Problem

Spotify’s current Terms of Use allow users to upload AI generated music, provided they hold full rights and avoid impersonation. The platform prohibits using its catalogue for external AI training but still doesn’t require artists to declare whether a track was made with AI tools.

That leaves a regulatory gap: AI generated music can be streamed, monetised, and added to playlists with no indication of its origin. Spotify’s policies protect the company’s data but not the creators whose catalogues may be used to feed AI models. The platform’s terms urgently need updating explicit labelling, new royalty categories, and rules that reflect the blurred reality of human-AI collaboration.

The Million-Dollar Question: Who Trains the Machines?

The biggest unresolved issue in AI generated music is training data. If models like Suno or Udio have been trained on copyrighted catalogues, then the resulting songs are built on unpaid creative labour. Artists’ recordings become digital raw material, stripped of consent and compensation.

The solution lies in royalties and transparency. AI developers should pay licensing fees or royalties when using protected works to train their models. Emerging academic work on “training-time attribution” could make this feasible, identifying which human-created material influenced a given AI output. Without it, the entire AI generated music industry risks collapsing under its own ethical contradictions.

Opinion: Act Before the Line Disappears

The music industry must draw a clear boundary before AI generated music overwhelms human artistry. Labelling, watermarking, and royalty reform aren’t optional -they’re essential. The public deserves to know when music was made by a human and when it was made by an algorithm trained on someone else’s unpaid creativity.

Music is not a dataset. It’s a reflection of human experience, shaped by emotion, struggle, and intent. Until AI companies acknowledge that by paying fair royalties and revealing their data sources, AI music will remain less an innovation than an appropriation. Ether the next generation of artists will be credited creators – or invisible datasets feeding machines.