A new global survey has revealed a striking development in music listening: nearly 97 % of respondents failed to distinguish between music composed by humans and tracks generated entirely by artificial intelligence. The survey, conducted across eight countries, underscores how quickly AI is reshaping the soundscape of the music industry.
Streaming platform Deezer partnered with research firm Ipsos to poll 9,000 people in the United States, United Kingdom, France, Brazil, Canada, Germany, Japan and the Netherlands. Participants listened to two AI-generated samples and one human-made track, then attempted to identify which was which. Surprisingly, the vast majority got it wrong. About 71 % of respondents expressed surprise at their inability to make the distinction. Deezer also reports that around one-third of the 50,000 new daily uploads to its platform are wholly AI-generated.
The findings raise serious implications for artists, copyright holders and streaming platforms. If AI-generated music is indistinguishable from human-made work for most listeners, questions arise around attribution, royalty structures and artistic authenticity. The scale of AI-generated uploads signals a potential shift in how music is created, distributed and monetised.
Deezer CEO Alexis Lanternier commented that the results highlight the urgent need for transparency in music creation and consumption. The platform has begun tagging tracks identified as fully AI-generated and is working to exclude suspected fake streams from royalty payouts. Meanwhile, labels like Universal Music Group are entering licensing deals with AI music companies to ensure that models are trained on legal content and proper rights frameworks are in place.
Key issues going forward include: how streaming platforms regulate uploads and royalty flow for AI-driven music; whether consumers will demand clearer labelling of AI-generated tracks; how rights holders respond to potential mass-scale production of AI music; and how human artists maintain an advantage when artificial creations sound so similar.
AI has matured to the point where most people cannot reliably tell if a song was composed by a machine or a human. For the music industry, it marks the transition of AI from experimental novelty into mainstream creation, forcing artists, platforms and rights-holders to rethink the very nature of musical authorship.