The Plight of the Music Industry in The Face of Generative AI
Image credit: Jason Leung on Unsplash.
The presence of Generative AI in everyday life has increased exponentially since its initial public release in the days of AI-Will Smith eating horrifically rendered spaghetti. I know for a fact that I’m not the only one meticulously looking through food delivery sites, trying to figure out if the cheese oozing from a burger I’ve been eyeing up is a quirk of the camera, or a sign that I’m seeing the accumulative sum of an algorithmic attempt to create the image of an average burger. Generative AI has, however, begun to be placed in creative industries as well as these more benign everyday examples. The adoption of Generative AI by music producing companies leaves many in the music industry concerned by the increasing prevalence of this new technology, with over 50% of artists, producers, performers, and writers claiming that they feel AI poses a significant threat to their career (This is Music 2025).
A surprisingly quick addition to the conversation surrounding Generative AI’s space in music is the addition of entirely AI-generated *musicians* and *bands*, creating simulated pop stars based off of AI-generated content. ‘Velvet Sundown’, for example, generated millions of listens, with their 1970s *inspired* indie-folk music, and even photos of the *band*, being entirely AI generated. When called out on this, those behind the image of the band maintained that the band was run by humans, until further pressure made them relent the involvement of Generative AI. This is becoming an increasingly common occurrence as Generative AI develops, with *artists* being generated across multiple genres, from American country music with ‘Breaking Rust’ and ‘Cain Walker’ themselves generating millions of listens, to R&B/Gospel music with ‘Xania Monet’ being the first AI-generated *artist* to enter a Billboard top 30 music chart (Billboard, 2025).
Though these examples demonstrate how AI has been quickly adding itself into the music scene since its inception, the drama surrounding ‘Velvet Sundown’ presents a significantly more insidious aspect of Generative AI’s space in the music scene: that there are currently no laws or legal frameworks requiring transparency of the use of Generative AI in music creation, even if the music was entirely AI-generated. In this regard, the lack of laws or legal frameworks restricting Generative AI’s space within the music industry demonstrates a significant lack of safeguarding for music artists, with an already precarious career made all the more unstable by this new technological paradigm.
When asked to comment, Martin Iddon, composer and Professor of Music and Aesthetics at the University of Leeds, summed it up when saying that a key problem of AI in the music industry is how “[listeners] feel duped, like they’ve been taken for fools, by an industry that also doesn’t always seem to get that one of the things that music fans really want is that sense of connection with artists (and artists who might have artistic intentions that you can sense in the music).” This is an issue that extends not just with listeners, but with radio producers, who are now placed in a position where they must examine every piece of music they’re given to put on-air with an increased measure of scrutiny, just to see if they’re unknowingly putting something generated by AI on their programme.
This may initially seem like speculation on a far-off future, but the effects of Generative AI in music are already being felt in the contemporary music scene. Spotify themselves revealed at the end of 2025 that they removed 75 million AI-generated songs in 2025 alone (Spotify Newsroom, 2025). For comparison, their current catalogue numbers around 175 million songs. One such example was an AI-generated *band* created as a replica of ‘King Gizzard & the Lizard Wizard’, who had left Spotify some months prior out of protest for their links with Palantir. Spotify quickly removed the AI-generated artist, but its existence drew many eyes toward the very real possibility of “headstrong” bands being replaced by AI-generated copies. Despite Spotify’s quick response to this, many well-known AI-generated *artists* such as ‘Xania Monet’, and even ‘The Velvet Sundown’, remain on the platform, making the specifics of Spotify’s stance somewhat confusing.
What ‘Velvet Sundown’ and other AI-Generated *artists* represent is a distinct uncertainty in the future of the music industry going forward. Warner’s deal with Suno in November of 2025, a year after their lawsuit against them for using their artists’ likeness in their AI programme (The Guardian, 2025), places artists in a position where their work can be used by anyone using Suno to create whatever they want (though at this stage, Warner’s deal with Suno remains something that artists must opt-in to for their work to be used by the programme). Warner has continued to keep this deal with Suno, despite backlash from artists and the general public.
What sets Generative AI apart from other introductions of technology into music production is its inherent sense of exploitation in its framework. Music based on Generative AI has been taken from data scraped from artists all across the internet, oftentimes based on contracts for public use before the issue of Generative AI became prevalent. When asked to comment, Samuel Murray, PhD and lecturer in Music Management at the University of Leeds, said “Machine Learning AI has been using music without seeking permission from rightsholders, vocal data has been scraped from songs and voices are being manipulated to do the humanly impossible.” As of right now, the music industry is faced with a developing technology nearly indistinguishable from human voice, with little to no protections placed for individual artists, prioritised by production labels looking to capitalise on a growing market.
While demonstrating the current scope of Generative AI’s impact on the music industry is important (and cathartic!), I don’t believe it’s enough to simply highlight the issue currently faced. AI, despite all of its frustrating qualities, is a technology that is unlikely to disappear anytime soon. It would also be disingenuous to claim that AI in the music industry is an entirely negative factor, as there are undoubtedly many ways in which it could be used to further music production without infringing on the rights or creative process of artists (though personally, this would also require major changes on the ethical ramifications of data extraction for AI training, as well as the environmental ramifications of AI, which are both topics outside the scope of this article!). When asked about this, Dr. James Mooney, Associate Professor of Music Technology at the University of Leeds, commented that “Like many new technologies, it has the potential to open up new creative possibilities when used in good faith. But it also has the potential to undermine musicians’ livelihoods, if used to generate musical content in a way that bypasses musicians.”
It’s the point of “good faith” that then becomes the sticking point. As it is right now, Generative AI lacks the proper restrictions on its data usage to properly protect artists’ rights. Similar to the early days of the internet, we are in a “Wild West” period when it comes to Generative AI: a new technological development we have no precedent for has suddenly come onto the playing field, and we’re scrambling to make sense of how to react. In this, then, lies our collective power. Pressing for more limitations on what data Generative AI has access to, how it can use it, a legal precedent for transparency in its use, and protections for artists against its unfair use, will decide how Generative AI is treated from here on not just in the music industry, but for all creative industries.
At this point, Generative AI’s role in the music industry is muddled, uncertain, and confusing. Yet, this is also the point where we can enact the greatest change in how we respond to it, and by extension, press legislative and legal responses to it. AI is here, and there’s no changing that now that the metaphorical Pandora’s Box is opened, but we can decide how we deal with the consequences, and press those in power to prioritise the safety and livelihood of the artists currently at risk.
Words by Connor Haresign