Will AI save the music industry or destroy it?

Will AI save the music industry or destroy it as we know it? In this guest post, Jacob Varghese, Founder and Director at Noctil, delves into the potential for AI to transform the music business while raising concerns about copyright infringement and fair compensation.

Will AI save the music industry or destroy it?

by Jacob Varghese, Founder and Director of music and movie industry data and cloud solutions provider Noctil, explores the double-edged sword of AI and music.

AI is the big talk of the town these days. Whether we like it or not, it is impacting nearly every aspect of our lives and the music industry is no exception. While there are debates in favor and against AI in all industries, it feels like arguments against AI are felt strongest in the music and creative industries. For the music industry specifically, while it offers immense potential to transform and improve the efficiencies of how the industry operates behind the scenes, there is a big question mark on how it will shape the future of music creation, copyright and consumption.

After all, in an age when we – as an industry – are putting in the effort to make sure rightsholders are compensated fairly, why should we accept this new technology that may put us back to where we were with illegal downloads over a decade ago?

One of the most pressing concerns surrounding AI in the music industry is the potential for copyright infringement. AI algorithms can generate music that closely resembles existing works, raising questions about ownership and originality. Rightsholders and artists rightfully fear that their creations could be exploited without proper compensation or attribution. This issue has already led to legal battles and calls for stricter regulations to protect intellectual property rights. In fact, earlier this year Tennessee became the first US state to pass legislation to protect musicians from unauthorized AI impersonation, while the RIAA and the major record labels filed a lawsuit against generative AI audio startups. 

While AI is highly contested on the “fake” impersonation and copyright infringement side, it also has the potential to be a powerful tool for creativity and innovation. AI-powered software can assist musicians in composing, arranging, and producing music, it even can help identify emerging trends and preferences, enabling artists to tailor their music to specific audiences. 

“it’s just not the futuristic, dystopian type that we see in scaremongering articles”

On the consumer side, AI has been in action long before the words “generative AI”, “OpenAI” and “ChatGPT” made headlines. AI and data science has been in use for a long time – it’s just not the futuristic, dystopian type that we see in scaremongering articles. Algorithms – or “recommendation engines” – that recommend new music, and movies, display relevant ads on social media, or an Uber nearby – these are all powered by algorithms that are based on data science. Through AI, DSP platforms can improve the user experience and match their listening habits with new, undiscovered music. DSPs analyse audio features, metadata, and user interactions to recommend songs or artists matching a particular mood, tempo, or style and offer personalized recommendations. 

Arguably, one area where AI could do the most good for the music industry is one that is hardly noticeable to creators and audiences. That area is the industry’s supply chain. Why? First, let me explain that I’m not talking about the supply chain in the sense of the distribution of music from a record label to physical and digital stores, and DSPs (although it is related). We’re talking of the supply chain in terms of the data and metadata that is associated with music works. These works travel in various directions (e.g. radio, DSPs, physical sales, sync) and the challenge is to match the use of these works with the work itself and then with its rightsholder. If these matches don’t take place correctly and efficiently, the royalties to artists and rightsholders will be either delayed or never arrive. 

But hey, what does AI got to do with this?! 

I’m getting there! So, behind every track, there is data – metadata – that can identify a track and match it to its rightsholders. It is not just the name of the song, artist and album. It goes into far much more detail. This includes songwriters, producers, publisher(s), record label(s), copyright holder(s) and of paramount importance: essential identifiers like ISRCs, ISWCs, and IPI numbers. And because there is no international standard (yet) a lot of this information is missing or gets entered incorrectly as it moves between various databases. Each database can have a different format of information which causes the various works out there in the supply chain to not communicate with each other. 

Enter AI. 

The good thing about AI is that it can handle vast amounts of data. Record labels, publishers, collecting societies and other music administrators handle hundreds, if not thousands, of lines of data on a regular basis. AI can make metadata management much easier and remove the human error that can be so easily introduced in our daily use of spreadsheets. AI algorithms can improve the accuracy and consistency of metadata, reducing errors and streamlining the royalty distribution process. A higher metadata matching percentage leads to a more equitable distribution of revenue among rightsholders.

It can go even further – machine learning can also be harnessed to identify patterns in data, improve metadata matching, de-duplicate, and even detect anomalies that might indicate lost royalties or fraud. All of this can have an impact on rightsholder royalties, the amount artists get paid and the speed with which they receive their money. On the business side, it can also help reduce operational costs and hours lost tracking data in spreadsheets, helping them focus on other important aspects of their jobs. 

Industry Collaboration

“finding a delicate balance between fostering innovation and safeguarding artists’ rights”

AI could be the closest to a solution to the metadata problem. But to get rid of the problem entirely, the industry needs to work together. Similarly, the industry needs to come together when it comes to AI. The key to harnessing all of the aforementioned AI benefits while mitigating its risks lies in finding a delicate balance between fostering innovation and safeguarding artists’ rights. This requires collaboration between industry stakeholders and policymakers.

To ensure that AI is used ethically and responsibly in the music industry, all stakeholders must work together to establish guidelines and standards. This includes developing clear definitions of AI-generated content, ensuring transparency in the use of AI, and protecting the rights of artists and creators. Tech companies that create AI-powered systems are also responsible for adhering to guidelines and standards.

The industry can’t wait for the law to catch up and risk facing another Napster moment.