Universal Music Group has placed itself at the forefront of the music industry’s current engagement with AI technologies: licensing deals with startups, copyright infringement lawsuits and lobbying campaigns all included.

As the company’s SVP, strategic technology, Chris Horton is at the centre of UMG’s AI moves, and will be appearing at our Music Ally Connect conference in January as part of our AI and licensing panel. Ahead of that, we interviewed him about some of the hot topics.

Music Ally: Please tell us a bit about your role at UMG as the Head of UMG’s AI review and task force teams. What does this entail?

Horton: First, it might help to have some background.  Just as we have historically embraced other technologies, we’ve been using AI in various ways for years, and even have a few patents and a team building AI tools for internal use.  Generative AI creates a new set of opportunities for artists, fans, and employees, but it does have challenges, and that led to creation of the AI review team.

The goal of the review team is to evaluate potential generative-AI related projects and guide employees to ethically trained systems while avoiding common AI-related pitfalls.

“Ethically trained” means trained on content that is licensed, owned by the platform, or in the public domain, rather than trained on copyrighted content that is used without the rightsholders’ authorization. 

An example of an AI pitfall is copyright eligibility.  In jurisdictions like the US, simple prompt-generated content is not eligible for copyright protection, which we need for our songs, cover art, etc.  We therefore direct people away from AI-generated images for cover art, even if the tools are ethically trained.     

The AI review team’s work includes managing incoming requests from labels and other internal groups about projects involving generative AI, researching whether the associated tools are ethically trained (shout out to FairlyTrained.org), and helping requestors avoid other potential pitfalls like lack of copyright. 

If we have to direct people away from certain tools, we try to find ethically trained alternatives.  We’ve reviewed hundreds of requests, and many people have been working on this behind the scenes.

The task force’s primary goals  are information-sharing and assuring that relevant employees across the company are aware of developments in the world of AI, as well as coordinating various internal projects and support groups.

Music Ally: You were involved in the first successful download and subscription services for UMG and you’ve seen a lot of different technologies over the years, including high-resolution audio, NFTs & web3 and AR and VR projects. Where do you think generative AI sits within all these developments? Just how significant do you think generative AI will prove to be for the industry?

Horton: Generative AI will be very impactful, up there with the smartphone.  I expect we will see new ways to create music (beyond prompts), new types of music, and entirely new music products/services we have not conceived of yet.  There will be new ways to engage with the music and artists we love.  There are some exciting product concepts in development, and it still feels like early days.

Music Ally: UMG has been at the forefront of some of the early deals with generative AI companies. What are you looking for in a partner here, beyond just the willingness to seek licenses in the first place?

Horton: We’re looking for companies with ethically trained systems that can also take an artist-centric perspective and reflect that in their products and services. 

Ultimately we exist to help artists achieve their greatest creative and commercial potential and the willingness of UMG and our artists to explore an opportunity will depend on the product and business model.  For example, text-to-music platforms that generate content and serve as a replacement for artists are obviously going to be a tough sell. 

Artists are more inclined to support platforms that help them express their artistry, improve their connection with fans, or have some clear benefit that will get them on-board.

Music Ally: One of the biggest challenges for licensing AI lies in being able to accurately attribute the output of such services in order to be able to pay the relevant writers, artists and rightsowners. How confident are you that technology will be able to help to solve this problem and are there any particular developments in this space you would highlight.

Horton: There are several companies who have built or are working on attribution systems, and there are several different approaches, all of which seem promising.  I don’t think we need perfection to move ahead, we just need industry participants to agree a given approach is sufficient. 

Furthermore, several of the generative music companies have attribution as part of their underlying architecture.  Attribution will continue to evolve in the near future, but we already have viable solutions for at least some products.

Music Ally: Here in the UK we’ve seen bodies representing musicians and songwriters calling for rightsholders to seek their permission before signing any licensing deals. What is UMG’s policy on this?

Horton: We are in the process of reaching out to artists to understand their perspectives, which will lead to getting buy-in for generative AI product categories that we think will be beneficial, supported by business models that we expect will generate meaningful revenue over the long term. 

Music Ally: Looking ahead a few years from now, what do you think we will be talking about when it comes to AI and music? How do you see this space developing?

Horton: This is the most exciting part to me.  Development will continue at pace, but I hope that, just like in the P2P era, licensing, litigation, and legislation will eventually align so that all AI companies see that cooperation with creative industries is the obvious path forward.  This will unlock new investments and businesses, leading to even more products and services. 

Long term, artists will get closer to fans, fans will have new ways to experience music.  Music will be more interactive and more responsive, but not all the time, because we still want shared experiences. 

There will be life-altering developments in music and health, facilitated by AI, new genres of music, new kinds of interactive visual and spatial music experiences, new instruments and creation tools.  And musicians will find incredible uses for AI systems we can’t predict today.  

So hopefully we’ll be talking about a combination of products we saw coming, as well as products and companies that surprised us, in a good way, and how “future now” is the best time to be an artist and a music fan.

Music Ally Connect 2025 takes place on 20 and 21 January in London. You can buy tickets here.