By Ivy Nicholls, Third Year, Liberal Arts
What is music without a story, a human voice or a personality? It’s hard to conceptualise this question until you’ve had the jolt of hearing a song and thinking, ‘is this AI?’ It’s a feeling of deception - the person you thought you were engaging with is a patchwork of data in the shape of a song, barely hidden like three raccoons in a trench coat. According to streaming platform Deezer, only 5% of British users can correctly identify artificial tracks - if you haven’t ever questioned this, you may be in the majority.
Nonetheless, artificial intelligence was detected in 40% of music released in July 2026. As it becomes both inescapable and harder to detect, the music industry faces new questions regarding authorship. We are also forced to confront the role of a listener and how technology has shaped it.
The most popular AI music spans the bland and the comedic. Many have the uncanny familiarity of the out-of-copyright music played at a Mcdonald's at 2am; others are built on absurdity and irreverence in the spirit of meme culture.
They are shortcuts to crudely rendered atmosphere or cheap laughs. Sometimes the line between the two is lost to internet post-irony, leaving me wondering what the intention is. What is music if it just exists to amass streams?
The popularity of AI music reflects a music industry that serves the passive mind, the backgrounds of videos and cafes, leaving the most lucrative gaps in the spaces between thinking. Where generations of convoluted, misremembered nostalgia make complete derivation the final goal. As critic David Quantick prophesied in 1986, ‘pop will eat itself’, obsessed with its own past. The algorithmic optimisation of this process leaves the only opening for the computers themselves.
Platforms are also obscuring the context of music. While Tiktok has permitted the popularity of older tracks for a younger audience, they are detached from their history. On Spotify, the words ‘britpop’ and ‘madchester’ are attached to AI-generated playlists that barely brush the 90s genres they reference. Chronology is disregarded in favour of vibes, moods and atmosphere. People and stories are obscured leaving a gap that AI can fill in- and it already is.
When all of music history is flattened into data, time is elastic. Michael Jackson can be resuscitated to perform a country version of ‘Billie Jean’ to 1.6 million views. Genres can be switched with the click of a button on Suno, and 50 Cent can perform ‘In Da Club’ from a 1950s soul club. I assumed that these novelty tracks fell under the comedic umbrella, but many comments express their support in earnest, some even wishing ‘he sung it like that.’
It may be harmless fun, but it makes me wonder if children with internet access understand the difference between these hallucinations and the work of the artists they are parodying. What if both end up being valued equally? When a song pretending to be by Eminem collects more views than some of his actual work, this could be reality. This is a landscape in culture as well as politics where we are not just post-truth but post-logic, the means of creation holding no bearing on a song’s meaning. Nothing has to make sense.
Lucy Bevan
This projection may be far-fetched. The influx of AI music in the last few years is not without resistance, Australia banning it from their charts and Spotify labelling AI artists and blocking them from personalised recommendations. It may disappear into obscurity once well-regulated, but it is important to consider how this absurd phenomenon has been made possible.
It doesn’t surprise me that the force that assesses and creates demand through streaming platforms has the power to supply as well. However, in a world of AI slop, human innovation holds a premium.
Featured image: Photo by Aleksandr Popov / Unsplash
Would you ever listen to AI music?
