Qobuz Reveals 60% of Its Identified AI Music Streams Are Flagged as Fraudulent

Qobuz starts labelling AI music as it reveals 60% of identified AI track streams are flagged as fraudulent.

Share
Qobuz Reveals 60% of Its Identified AI Music Streams Are Flagged as Fraudulent

Music streaming service Qobuz has started visibly labelling AI-generated music, while revealing striking figures about how much of that content is actually being heard.

The French high-quality streaming and download platform now displays a tag when music has been identified by its systems as AI-generated.

Qobuz says 60% of streams from tracks identified as AI-generated are currently deemed fraudulent by its anti-fraud systems, meaning those streams are excluded from reporting and royalty payments.

The numbers become even more striking elsewhere. More than a third of albums identified as AI-generated received no listening activity, covering around four million tracks. Qobuz says those tracks have been removed from its search engine.

Meanwhile, just 0.38% of all streams on Qobuz currently come from tracks identified as AI-generated. AI-generated material is also excluded from Qobuz's human-curated selections and algorithmic recommendations.

However, Qobuz isn't rejecting every use of AI in music. Its AI Charter supports artists using artificial intelligence for purposes including mixing, mastering and assistance with composition while opposing mass-produced fully AI-generated content designed to saturate catalogues or manipulate royalties.

The company also acknowledges an important limitation: detecting AI-generated music isn't perfect, particularly when human and AI creation have been combined.

For listeners, the new label means they can now see when Qobuz believes the music they're playing was generated using AI rather than discovering that information elsewhere.