Introduction

The debates around AI remain at the top of the agenda for anyone involved in the media and technology sectors. Following on from our panel discussion on AI and the music industry at SXSW London, we look at some of the emerging themes in an industry at the forefront of technological innovation and disruption.

Major labels and AI partnerships

Just a few years ago, the relationship between major labels and AI developers was most obviously characterised by a wave of lawsuits being brought by rightsholders, including  against developers such as Suno and Udio.

While several of those lawsuits remain ongoing, the last few years have also seen some settlements, as well as increasing areas of cooperation between the music industry and AI developers. The most recent of these was Spotify and UMG’s announcement of “landmark recorded music and music publishing licensing agreements enabling Spotify to launch a new tool allowing fans to create covers and remixes of their favourite songs”.  While the specific details of these deals will no doubt remain an area of interest and heavy scrutiny, these deals show an evolving licensing market and an industry that is experimenting with a variety of new products and revenue streams. The evolving nature of this licensing market was just one of the matters addressed in the UK Government’s recent report on copyright and AI, discussed below.

The Government’s AI and Copyright Consultation

The UK Government has recognised its creative industries and AI sector as twin pillars of economic growth. Copyright sits at the heart of the creative economy, providing creators with the means to control how their works are used and to seek payment for that use. However, the rapid expansion of the AI sector has placed considerable strain on this framework: the widespread use of copyright-protected material to train AI models has presented rightsholders with new and complex challenges. Balancing its ambition to grow the UK's AI sector — as set out in the AI Opportunities Action Plan — with its commitment to supporting the creative industries, the government took the view that the existing legal framework required review.

It is against this backdrop - a legal landscape that left both the ownership and protection of AI outputs and the lawfulness of AI training in a state of commercial uncertainty - that the UK Government published its consultation on AI and copyright on 17 December 2024. The consultation set out four possible approaches: 

• no change to copyright and related laws;
• strengthening copyright protection through mandatory licensing;
• introducing a broad TDM exception without rightsholder controls; or
• creating a TDM exception coupled with opt-out and transparency mechanisms. 

The government's preferred direction was to broaden the TDM exception to permit commercial AI training by default, supplemented by an opt-out mechanism for rightsholders and enhanced transparency requirements on model developers. This preferred position, and the pace at which the government appeared willing to move, prompted significant engagement from across the creative industries.

Creators’ Resistance

Industry organisations, including the Ivors Academy and the Musicians' Union, argued that the proposals risked enabling the use of creators' works for training without permission, transparency or fair payment. UK Music and the British Phonographic Industry (BPI) focused on what they characterised as the systemic commercial risk: in their view, a statutory exception would legalise large-scale, uncompensated use of protected content at the precise moment a commercial licensing market was beginning to emerge, thereby reducing the incentive for AI developers to negotiate voluntary deals.

Arguably the most visible moment of public mobilisation came in May 2025, when more than 400 UK artists, including Paul McCartney, Elton John and Dua Lipa, signed an open letter coordinated by Baroness Beeban Kidron, urging  the  Prime Minister  to ensure that creators retained meaningful control over how their work is used in AI development.  

Following this sustained public, parliamentary and sector-wide engagement, the government changed tack. In its report - the Impact Assessment on Copyright and Artificial Intelligence, published on 18 March 2026 - the government concluded that it would not introduce copyright reform until it was confident that any changes met its objectives for both the economy and UK citizens. 

Looking to the future

Throughout history, leaps in music technology have led to evolutions in both the creative process and the way that music is consumed. Without the invention of the electric guitar, the synthesiser, and sampling, much of modern music would be wildly different. Without the invention of the CD, or streaming, the way that music is consumed would be wildly different. Is AI yet another evolution that progresses the way that music is created and consumed, or is it something more revolutionary? What is clear is that the way in which music is created, monetised, and protected will inevitably continue to evolve - and AI is playing its part in that. So, how much will AI change the music industry, and what is the likely direction of travel for the future? 

1) Streaming platforms are likely to continue developing systems to navigate the nuances around AI labelling and the challenges surrounding royalties and streaming revenue.

Breaking Rust, a fully AI-generated country and outlaw-blues music project, topped Spotify’s US Viral 50 with Walk My Walk and Livin’ on Borrowed Time in November 2025, and Walk My Walk led Billboard’s Country Digital Song Sales chart for three consecutive weeks. This project's commercial success highlights the growing capability of AI-generated music to compete directly with human-created works for listening time, and underscores the urgency of the unresolved questions around labelling, licensing and royalty distribution. The streaming industry has been grappling with whether to label AI-generated music, and how to do so given the significant nuances of how AI can be used during the creative process.

The use of AI in both generating and enhancing music is also becoming more sophisticated. A 2024 YouGov survey found that only one in five Americans were confident they could spot the use of AI in music, and in a 2025 controlled Deezer–Ipsos poll, 97% of listeners failed to correctly distinguish between AI-generated and human-made tracks. 

The UK Government currently has no plans to mandate AI labelling for music; however, platforms have already started taking measures to do so: 

• Deezer has taken the most proactive stance - once a track is determined to be "100% AI-generated," Deezer flags the host album with a visible "AI-generated content" label and removes the track from all algorithmic and editorial playlists to assist with royalties being distributed to human artists.
• Spotify, by contrast, has taken the position that the use of AI in music creation is not binary and accordingly rejects the view that the industry should classify every song as either "AI" or "not AI". Spotify has noted that artists and producers may choose to use AI to help with some parts of their productions and not others, and has expressed the view that disincentivising listening through labelling or removing tracks from algorithms could unfairly prejudice artists that use AI as part of their creative process.
• Apple Music, meanwhile, announced the introduction of "transparency tags" and indicated it will eventually require labels and distributors to self-disclose AI involvement in new releases.  

While output transparency is an important factor in the increasing prevalence of AI-generated content on platforms, debates around the appropriate standard are likely to continue. 

2) It is likely that artists and labels will look for new ways to monetise their work and engage with their listeners.  

Artists are by their very nature innovative, and have utilised previous developments in technology to find new ways to engage with their fans and monetise their work. For example, Amplify World, a tech platform co-founded by Kasabian drummer Ian Matthews, has provided over 94,000 artists worldwide with both data analytics and direct fan engagement tools, helping them to build visibility and develop sustainable income streams.

AI is likely to create possibilities for further methods of promoting engagement with listeners. Indeed, Spotify announced its new remix tool in May 2026, enabling Premium users to remix songs, as part of a licensing deal struck with Universal Music Group. The next year will likely see the development of further innovations and collaborations between rightsholders and AI companies (as well as other industry players). 

3) The utilisation of AI as part of the creative process is likely to become even more normalised – and can assist artists during the production and sound engineering process.  

Creators have already been using AI to assist with the production and sound engineering of their work, for example restorative work such as isolating vocals and layering them on top of different instrument channels. The London College of Contemporary Music cites AIVA and Amper Music as platforms that allow artists to use AI to assist with sampling and generating synthesised sounds to add to existing work.

In this sense, AI sits alongside sampling, drum machines and synthesisers as a tool that artists can use to create. It is notable that the genesis of sampling, drum machines and synthesisers in the 1970s and 1980s also created widespread alarm before being absorbed into mainstream production practice.

Conclusion

The music industry stands at a pivotal juncture. The rapid advancement of AI technology has tested the boundaries of existing legal frameworks, prompted significant commercial realignment among major industry participants, and catalysed an unprecedented level of cross-sector engagement on questions of copyright, consent and compensation. History suggests that the music industry has consistently demonstrated a capacity to absorb technological disruption — from the electric guitar to the synthesiser, from the CD to streaming — and to emerge with new creative possibilities and commercial models.