More AI-generated music has hit Spotify, and creators are concerned that this competes with authentic work.
We now have at least three major AI tools in the text-to-audio (or, perhaps more to the point, text-to-music) space: Udio, Suno, and Limewire.
While music production has become heavily democratized since the 1980s and 90s, anyone can now produce natural-sounding music in seconds.
From there, it just takes a few actions to upload it to streaming platforms and monetize.
Ex-Stability Audio team leader Ed Newton-Rex, who has now become a firm advocate for the proper licensing of data by generative AI companies, brought to attention a selection of obviously AI-generated songs that have racked up thousands of listens.
Spotify is actively recommending AI music made by AI models that may be trained on copyrighted music without permission. 240k monthly listeners for this ‘artist’. Half a million streams of the top track.
This eats into the royalty pool available to human musicians.
Generative… https://t.co/Z9semlqrO8
— Ed Newton-Rex (@ednewtonrex) April 29, 2024
It’s evidence of an existential challenge facing the music industry.
As Newton-Rex points out, AI-generated music created by tools trained on authentic music now competes with the very authentic music it was created from.
This mirrors a near-identical situation in the visual arts, where tools like MidJourney, trained on vast quantities of data without copyright holders’ permissions, now directly compete with artists in industries like graphic design.
Not everyone is concerned about this, however. Technological democratization has been key to art’s evolution, particularly in the last few centuries, dragging it from the purview of the upper classes to something wider society can create and engage with.
Some question also: what is ‘authentic’ creative work in our digital era, anyway?
If AI’s role in art is somewhat permitted, who draws the line when it becomes unacceptable?
Should people make indiscriminate moral judgments of creators who wield AI? Is this practical or useful?
For many, however, the crux is that AI companies have not paid for the data they used to create their models. That seems quite clear-cut, at least.
Spotify and other streaming companies will light the way
One of the simplest fixes here is de-monetizing or even de-platforming AI-generated music.
Major platforms like Spotify, record labels, and big artists hold the biggest influence over future trajectories. However, implementing Draconian measures to terminate AI’s use in music is unrealistic.
In late 2023, Spotify became the center of the AI music controversy when it reportedly removed tens of thousands of AI-generated songs uploaded by Boomy, an AI music generator.
The move came amid suspicions that bots were artificially inflating the play counts of these tracks, illustrating the challenge streaming services face in ensuring fair compensation for human creators while combating fake streams and bot-driven manipulation of royalty pools.
The incident was followed by a rare public interview in September 2023 with Spotify’s CEO, Daniel Ek, who went some way to clarify the platform’s stance on AI-generated music.
Ek stated that while Spotify would continue to host AI-generated content, it wouldn’t support tracks impersonating real artists without their consent.
This came after Spotify removed the AI-generated song “Heart on My Sleeve,” which featured the voices of Drake and The Weeknd without their permission. “Heart on My Sleeve” was later refused to be considered for a Grammy.
Ek identified three distinct applications of AI in music: tools that enhance music production, those that imitate real artists (which Spotify does not support), and a complex category where AI-created music is evidently influenced by real artists but doesn’t impersonate them directly.
Concerns about AI’s impact on the music industry have since continued to grow as tools like Suno and Udio improve text-to-audio technology.
In April 2024, over 200 prominent artists, including Billie Eilish, Nicki Minaj, Pearl Jam, R.E.M, Chase & Status, and Jon Bon Jovi, vowed to tackle AI music head-on, while the US SAG-AFTRA performer’s union negotiated with record labels for protection for artists from AI music.
Another subtext here is that these big artists will hold even more power in the AI music era.
They have the eyes and ears to spot AI music created with their assets, the financial clout to sue, and influence over record labels.
Some artists have already struck deals to monetize their musical assets to companies for AI tools. T-Pain, John Legend, and seven others entered a deal to enable YouTube to replicate their voices as a feature for Shorts “to shape the future of AI in music.”
One thing amongst all this seems certain – the small indie music producer with fewer bargaining chips will feel the full force of AI-generated music before anyone else.
