Copyright foundations of the music industry shaken by the rise of AI

World IP Day, which takes place on 26 April, is an annual opportunity to reflect on the pivotal role played by intellectual property (IP) rights in encouraging innovation in science, technology and the arts. Innovation like this boosts economic growth in countries worldwide and sustains a healthy cultural environment in which vibrant and diverse ideas can be creatively expressed, shared and enjoyed.

The focus of this year’s World IP Day,  music is an industry which continues to embrace transformative technology and this technology will be protected by various IP rights, including patents. However, copyright is the IP right mostly closely associated with music.

Copyright and the music industry

Copyright is essential to the music industry. It prevents the creative contributions of artists from being copied by others and ensures that creators are financially rewarded for their efforts. Without copyright, there would be little incentive for artists and other innovators in the music space to continue pushing the boundaries and exploring new creative avenues. The nature and impact of copyright in the music industry is explored in another of our articles published in celebration of World IP Day this year.

The impact of AI on the music industry

Over the last couple of years, there has been an explosion in the use of AI and the music industry, as many other sectors, has been heavily impacted by this with the introduction of some positive changes around how we indulge and express our musical passions. However, wider use of AI has also led to some fundamental questions around the precise nature and scope of some IP laws. For example, the patentability of AI has arisen in the UK in the context of musical innovations, with the Supreme Court soon to rule on the issue of whether Emotional Perception’s artificial neural network music recommendation tool is patentable. 

AI tools can be used to compose music, generate lyrics and produce tracks, as well as assist in mixing and mastering recordings, making the production process faster and more accessible. AI has also transformed how music is consumed. For example, Spotify’s AI DJ uses data points that track listening habits and how they change throughout the day or by location. AI DJ learns the user’s music tastes to curate a personalised listening experience for them. It also automates the process of tagging tracks so they can be categorised by content or mood. This engages users by introducing them to artists and genres that they may have been unaware of before.

Arguably, AI has also had an even more revolutionary impact in the global musical sphere, resulting in groundbreaking musical innovations which would never had been thought possible five years ago.

For example, the Beatles used AI to release their first new song in 45 years despite the deaths of two of the band members. AI was used to isolate John Lennon’s voice from an old, noisy demo tape and restore it so that it could be added to parts previously recorded by the late George Harrison with Ringo Starr and Paul McCartney adding their own contributions, enhancing an old, low quality recording to save music from being lost.

Music generating technology like IBM Watson Beat, Amper Music and Google Magenta’s NSynth Super can enable those who want to create music content but do not know how to play particular instruments to create music with the sound of those instruments. The user is able to communicate a melody or chord and the AI tool then replicates the sound of the selected instruments or even generates entirely new sounds. The user can also ask the tool to generate musical output based on the type of music they may want to produce and the user can edit the output so that it fits with lyrics or vocals. This is a good example of output enhancing musical collaboration between human and AI.

The rise of generative AI

The creative sector, including the music industry, has felt the impact of the rise of generative AI (GenAI) most acutely. Most of us have spent at least some time engaging with GenAI to create AI generated output from text to images as well as tools that can be used to generate music from our preferred musical genre. In cases of purely private use, this may be little more than harmless enjoyment. However, when GenAI is deployed for commercial purposes this can have a serious impact on the balance of rights and rewards in the music industry. Creative content might be used without the consent of the copyright owner to train, or otherwise be utilised by, these AI tools particularly when the bargaining power of a major label is absent. This use can be entirely without the awareness of artists or other rights holders, and even if they are aware they may struggle to secure royalties they would otherwise be entitled to in return for the exploitation of their work. Artists who depend on royalties for their livelihood are disincentivised from artistic endeavour and, over time, this could have a devastating effect on the rich music landscape that we currently enjoy with musical content being released in much more restrictive ways.

In an echo of the Getty Images – Stability AI case, Universal Music Group, Sony Music and Warner Records have all sued AI music synthesis companies Udio and Suno in the US for allegedly committing mass copyright infringement by using recordings to train music-generating AI models, with the record labels seeking damages of $150,000 per song used in training the AI. As with the Getty Images case, this suggests that significant IP rights holders are taking robust action against the use of their content to train AI models by AI developers.

Also in the US, a music producer has been accused of creating hundreds of thousands of songs using AI and then using ‘bot’ accounts to stream the songs on platforms including Amazon Music, Apple Music, Spotify and YouTube Music. He allegedly used AI to create songs and then randomly generate song and artist names for the audio files so that they appeared to come from real artists. In a first of its kind prosecution, it is alleged that by doing this he fraudulently obtained over $10 million in royalty payments from the platforms.

AI models are also reportedly being trained using vocal snippets to create ‘deepfake’ songs that mimic the vocals or styles of famous artists. ”Heart on My Sleeve”, a track featuring vocal clones of Drake and The Weeknd, drew around 15 million views on TikTok and hundreds of thousands of plays across music streaming platforms in less than 48 hours.

The site ‘Jammable’ allows users to incorporate their own vocals with those of iconic artists, with a library of over 3,000 voice clones of well-known artists for users to choose from.  

Perhaps most notably is the fact that the rapid evolution of AI has the potential to shake the foundations of copyright as we know it, posing a direct challenge to the protection afforded to artists by copyright. AI development depends entirely on the availability of high quality training data and governments worldwide have struggled to strike the right balance between striving to be the jurisdiction which most favours AI development whilst recognising the right of creators, including musicians, to have their content protected from unauthorised use, and to be fairly rewarded where the use is authorised. We have previously written in more detail about these issues.

Striking the balance

The UK is just one such jurisdiction seeking to achieve a balance between AI and creative content. In December 2024, the UK government issued a consultation seeking stakeholder views on the best way to do that. The consultation closed in February 2025 and we are awaiting publication of stakeholder responses, and confirmation from the government on its preferred way forward. In the consultation document, the government expressed its initial preference for free use of copyright protected material by AI developers with the ability for creative content owners to opt their material out.  The music industry has been one of the most vocal critics of this approach with high profile artists, including Sir Paul McCartney, openly expressing their concerns for the future of the music sector should this approach be adopted.

Protection for AI generated works

On the other end of the AI process – output rather than input – the question of whether AI-generated content itself qualifies for copyright protection and  Currently in the UK, works generated by “computer” are protected by copyright, with the owner of the copyright being the person who makes the arrangements necessary for the creation of the work. In its consultation earlier this year, the UK government proposed removing copyright protection for computer generated works on the basis that they cannot satisfy the originality test necessary for copyright to subsist – because its essential focus is human intellectual originality – and in recognition of the broader ethical issue that content which is truly of human creation is more worthy of protection than that which is derived from machine. Stakeholder views were invited on this question, but responses are yet to be made public.

What next?

It is fair to say that the copyright foundations of the music industry have been shaken by the rise of AI and there remain many issues to be resolved by the courts but most likely by legislation. In the UK, the outcome of the pivotal AI v copyright consultation is awaited. The purpose of this article is not to argue for the cause of the music industry and against AI innovation – or indeed vice versa – but  simply to encourage anyone reading this article to take some time to think about the legal and ethical issues that arise. These are, after all, issues which ultimately will affect us all, music lovers or not.

Alicia Dyson of Pinsent Masons contributed to this article.