Velvet Sundown is a band that very soon found popularity on Spotify but has now admitted to being an AI creation. The group, styled to resemble an indie rock band with over a million monthly listeners, drew attention for its music, yes, but also for the mystery surrounding its identity. The band’s profile, recently updated, now confirms that the characters, voices, lyrics and even the music itself were generated using AI tools.
Social media posts were suspicious at conflicting backstories, while photos of the band members looked computer-generated. All of this had people talking online, with many unsure whether the band was real or part of a digital experiment. On Tuesday, AFP reported that Velvet Sundown described the project as an “artistic provocation,” pushing the boundaries of what audiences accept as authentic music.
Deezer, one of the streaming services that hosts the band’s music, has labelled Velvet Sundown’s songs as possibly generated by AI. Meanwhile, Spotify, where the group gained most of its popularity, declined to speak directly about the band. A spokesperson for Spotify said that the platform does not promote AI-generated music and that uploads come from licensed sources.
How Big Is AI Music Getting?
AI-generated music is becoming a big thing on digital platforms very fast.. In April, Deezer said that around 20,000 fully AI-created tracks were being uploaded to its service every day. That accounts for 18% of all daily uploads. Back in January, it was 10%.
According to CISAC, which represents over 5 million creators, musicians could lose over 20% of their earnings in the next 4 years due to this rise of AI music. In financial terms, that loss could reach around €4 billion by 2028.
This growth has started stressing artists out across the music scene. The main worry, other than AI replacing artists, is how easily fake content can flood platforms. Now, even though many of these tracks are never actually streamed, they fill the system up and open doors for abuse, including fake plays and copyright fraud.
More from News
- How Meta Is Helping The UK Government With Public Services
- Recent Studies Show The Mental Health Risks Of AI Therapy Choices
- What Influences Bitcoin Prices And Fluctuations?
- World’s First AI Chef To Come This September. Here’s How It Works
- Microsoft Who? Nvidia Has Officially Become the First Company to Surpass a $4 Trillion Market Cap
- Are AI Startups Investing In Teachers Learning AI A Good Move For Education?
- Can a Robot Really Perform Surgery Without Human Help?
- Reports Show Fewer Students Chasing Tech Careers, Here’s Why
Can AI Music Be Detected?
Deezer has developed its own detection tool to find AI-generated songs. This technology has in development for over a year and can identify music made by well-known AI models such as Suno and Udio.
According to Deezer’s Chief Innovation Officer Aurelien Herault, most tools today can detect AI music only if trained on data from a particular model. The new tool, however, works across models and does not need specialised training data.
Two patents have already been filed to secure the technology. These patents cover methods used to find digital fingerprints that separate synthetic audio from real human recordings. Deezer now plans to build on this tool by tagging AI music and removing it from recommendation algorithms.
CEO Alexis Lanternier said that although AI can be helpful for creativity, it must be used responsibly to protect the people who create music for a living.
What Do Music Fans Think About AI Songs?
A recent survey of 1,750 music fans in the UK, carried out by the BPI and AudienceNet, shows that people still care deeply about human creativity. More than 82% of those surveyed said human input is essential to music. Over 80% said they prefer human made music, and about 64% said they do not want to hear AI made songs that try to sound like their favourite artists.
The same survey found that 81.5% of people think AI music should be clearly marked. Another 78.5% believe that AI should not use an artist’s voice or music without permission. Over 72% want AI tools to list exactly what music they have used during training.
These views are global, and don’t end in the UK… In 2023, a global survey by the IFPI found that 79% of listeners across 26 countries felt that human creativity matters in music. More than 75% believed that artists’ work should not be copied by AI without consent.