As has been widely reported over the last few days, legal action is being taken by the Recording Industry Association of America against Suno and Udio, two startups which are offering generative AI music services which can be used to create new music, but whose critics argue that the material used to train the AI exploits the recorded works of artists in a way which constitutes a breach of copyright of “unimaginable scale”. In this article we examine the issues and their context, and ask for the experience of our audience.
There are some nuanced arguments running both ways here based around copyright ‘Fair Use’. The lawsuit alleges that Suno and Udio’s software unlawfully harvests copyrighted music to generate similar work and asks for compensation of $150,000 per work.
This video from a post on X shows how closely the output of this type of technology can mimic existing copyrighted recordings. The argument around Fair Use is investigated in our article Is Training AI Music Generation Apps With Copyrighted Music Fair Use?
What Is ‘Fair Use’?
Fair use is a crucial aspect of copyright law that allows for the limited use of copyrighted material without permission from the copyright owner. It serves as an essential safeguard for creativity, scholarship, and free expression. Fair Use enables individuals to use copyrighted works for purposes such as criticism, commentary, news reporting, teaching, and research. Central to this is the idea that such use is ‘transformative’, a test which AI music generation superficially seems to meet. The problem in this case is that these examples, and potentially in unlimited number of future ones, clearly reference specific copyrighted works and as such probably don’t meet this cirterion of being ‘transformative’.
Why Are The Labels Taking Action?
Generative AI clearly represents a threat to composers, songwriters, musicians and people in technical roles. But why sue over this particlular case? How is this different?
If we look at the timeline of technological threats faced by the music business we can see this is yet another in a long line of crises facing the industry:
Home Taping - Readers in the UK might remember how the cassette recorder was ‘Killing Music’ - a campaign instigated by the BPI to appeal to listeners’ morals not to copy records to tape. The BPI also proposed a levy on sales of blank media to reimburse labels for losses incurred from home taping.
Digital Recording - CD duplication at home via CDR drivers well and truly let the genie out of the digital duplication bottle but the introduction of consumer digital tapes like DAT and DCC and MiniDisc brought us the Serial Copy Management System. SMCS was a code which limited digital copies to a single generation only.
Napster - Digital duplication wasn’t welcomed but copying was limited by physical media, meaning you had to at least have access to a physical copy to duplicate it. Not so with Napster and its follow-on products like Limewire and various bittorrent sites which enabled peer to peer sharing over the internet.
Spotify and other legal streaming - Supposedly the answer to peer to peer music piracy was to ‘go legal’. By adding the convenience and reliability illegal piracy lacked, Spotify and others made most illegitimate music sharing irrelevant. But while this might have seemed a solution, who was it a solution for? the artist or the labels, or just the streaming platforms?
The one thing which strikes us about this potted history of ‘alternative’ music distribution is that the labels fought back to make sure they got paid, but artists always came out worst of all.
How Is Generative AI Different?
The thing which is different this time is that this isn’t about recordings. This is about songs. A specific recording has an owner. The songwriter or artist often doesn’t own that recording but if AI can generate new, original but derivative music which copies their style and delivery then that goes far beyond a recording they have made and threatens the value of recordings they might make in the future by harvesting the sum of all the recordings they have previously made. That’s different!
What Is Being Done About Generative AI And Music?
Apart from specific legal action like that reported here there is more positive action being taken to establish guiding principles around the use of generative AI in music.
The Seven Principles for Music Creation with AI were proposed by Roland and Universal Music Group. They are as follows:
We believe music is central to humanity.
We believe humanity and music are inseparable.
We believe that technology has long supported human artistic expression, and applied sustainably, AI will amplify human creativity.
We believe that human-created works must be respected and protected.
We believe that transparency is essential to responsible and trustworthy AI.
We believe the perspectives of music artists, songwriters, and other creators must be sought after and respected.
We are proud to help bring music to life.
Laudable aims indeed. For a more specific take on the same issue UK Music published their 5 key principles on AI:
Creators’ choice: The creator, or their chosen rights holder, should be able to decide if and how they want to use their creative talent. This certainty underpinned by legal rights (copyright) should not be undermined by any exception to copyright or compulsory licensing during the input stage. Users need to respect creator’s choice as baseline for any discussions.
Record keeping: It is important that in the input stage, the tech providers keep an auditable record of the music ingested before the algorithm generates new music. This is the only point in the process when these data points can be documented.
Human creativity: Without human creativity there should be no copyright.
Labelling: Music generated by AI should be labelled as such.
Protection of personality rights: A new personality right should be created to protect the personality/image of songwriters and artists.
So the problems are well understood, but are labels working with or independently of artists when trying to address this new development? Take our short survey and tell us your experience.
Have You Been Consulted?
Call us cynics but we suspect we might already known the answer to this question, but if you are a signed artist or have a relationship with a label please share you experience us with using our short survey, or share you thoughts in the comments.