Last week, I read an investigation by The Atlantic that exposed datasets containing more than 21 million songs used in AI music training, many taken from musicians who had no knowledge of, or control over, how their work would be used. Among them, I was shocked to discover songs I had spent months relentlessly labouring over five years ago, now Frankensteined with other artists’ music within the same ecosystem that helps platforms such as Suno turn musicians’ work into commercial product.
In a 2025 interview on the 20VC podcast, Suno CEO Mikey Shulman insisted that "it seems crazy that music should not be as engaging as Fortnite," elaborating that "people don't really enjoy the process of making [it]." Shulman's proposition to make one of the oldest creative practices in human history "as engaging as Fortnite" is to remove the creation process entirely from that practice. He suggests replacing human creativity, insight, and emotion with a machine that attempts to make sense of the intrinsic humanity of its training data, then mimics it. How well does the machine do it? Well, that depends on who you ask.
Towards the end of last year, the track 'I Run' by HAVEN. went viral on TikTok. Produced in a UK garage style, the track featured an uncredited female-sounding vocal that prompted online speculation due to its eerie similarity to Jorja Smith’s voice, despite the BRIT Award-winning singer denying any involvement.
Harrison Walker and Jacob Donaghue, the two producers behind the track, later confirmed that they had used Suno's AI-assisted vocal processing to re-synthesise their voices as a "soulful vocal sample". They denied allegations that they were trying to deepfake Smith's voice, whilst simultaneously using tags like #jorjasmith in now-deleted social media promotion of the song. The song was soon removed from major streaming services, and blacklisted from the charts following intervention from Smith's team. It was then later re-released with vocals from Kaitlin Aragon, reaching No. 9 on the UK Singles Chart and topping the Official Dance Singles Chart.
Comments such as "I don't care if it's AI, it sounds like 2015/2016. I like it" and "AI or not, this is a serious tune" can be found all over the original version's YouTube page, which has over 18 million views. Clearly, despite Smith's discomfort with her likeness being used without consent, many consumers will still listen to and engage with AI-generated music.
Some artists are less opposed to AI, though, with four-time Grammy-winning producer Timbaland starting his own AI entertainment company, Stage Zero, attempting to make the term "A-Pop" – artificial pop – catch on with AI/hybrid artists such as TaTa Taktumi.
Producer Diplo also positioned himself to be deeply pro-AI in an interview with podcaster Daniel Wall, claiming, "I don't even need a voice anymore. I can get the best voice from AI. I don't need anybody to sing the song." He elaborated on this further, stating, "There's no winning. There's no fighting AI […] If you are a creative, you need to adapt or just, like, give up and become an Uber driver until everyone has a Waymo […] I'm not gonna candy-coat the future."
Across each of these examples, there seems to be an underlying rejection of collaboration and community. The implied message is that, by utilising AI instead of collaborating, they don't have to compromise their artistic or commercial control over the song. AI won't protest in the studio, or put forward ideas and suggestions that may clash with a producer's initial plan for the song. Perhaps most importantly for these producers, though, is that it won't ask for royalties or fair monetary compensation.
That is what makes these examples so frustrating: each figure has access to a vast community of talented musicians, yet they choose to use AI despite its weaker emotional and sonic quality than that of a real artist. Just compare the vocal tones and timbres of TaTa Taktumi's 'Glitch x Pulse' to practically any vocalist Timbaland worked with in the 2000s, and the difference in energy and soul on the tracks is night-and-day.
The real joy I got from recording that collection of songs as a teenager did not come from sitting alone in my room recording endless guitar and vocal takes, but from the input of friends and fellow musicians – whether that was watching my friend who lived up the road play a Midwest emo-inspired guitar solo in one take, or another performing a delicate poem she'd written, or another arranging ornate layers of saxophone only to record them into GarageBand with apple earphones. Fundamentally, music is about building community and a sense of identity – the antithesis of what AI stands for. No matter how many songs it harvests, it will never be able to pull from the lived experience that makes a song truly resonate with the human soul.
George Barker
Image: 'Glitch x Pulse' Official Single Cover
If you enjoyed reading this article, please consider buying us a coffee. The money from this pot goes towards the ever-increasing yearly costs of running and hosting the site, and our "Writer Of The Month" cash prize.