Humans Can’t Tell Real Music from AI Anymore

A new global survey has revealed that most people can no longer tell the difference between songs created by humans and those generated by artificial intelligence.

The study, conducted by Ipsos for French streaming platform Deezer, asked 9,000 people from eight countries, including the United States, Britain, Germany, Japan, and France, to listen to short clips of music. Out of these clips, two were AI-generated, and one was made by humans. 

The results were shocking: 97 percent of listeners couldn’t tell which song was made by AI.

According to Deezer, more than half of the participants felt uneasy after realizing they couldn’t spot the difference. Many feared that AI could flood streaming platforms with low-quality content and reduce creativity in the music industry. 

About 51 percent said AI would bring more poor-quality music, while nearly two-thirds believed it could lead to a loss of originality.

Deezer’s CEO, Alexis Lanternier, said the survey shows that people still care deeply about authentic music and want to know whether a song is human-made or not. To address this, Deezer has begun labelling AI-generated music on its platform, the only major streaming service currently doing so.

The rise of AI in music has been rapid. In January, only one in ten tracks streamed daily on Deezer were AI-generated. Now, that number has jumped to nearly one in three, amounting to around 40,000 AI songs per day.

Eighty percent of those surveyed said they want AI-made tracks to be clearly marked. The demand for transparency has grown, especially after an AI-generated band called The Velvet Sundown went viral earlier this year. 

Their song gained over three million streams before it was revealed that no human musicians were behind it.

In response, platforms like Spotify have started encouraging artists and producers to disclose their use of AI in creating music.

As AI continues to blur the line between human creativity and machine generation, one thing is clear, listeners want honesty about what they’re hearing. The question now is: will the future of music be led by humans, or algorithms?