The greater potential of generative AI
Summary
AI tools and platforms are transforming the entire music production process, from composition to distribution, enhancing creativity while also raising concerns about copyright infringement.
As music consumption evolves with each technological advancement, from vinyl to streaming, AI is introducing a radical shift towards interactive and consumer-driven creation.
The rise of Gen Z’s participatory culture is driving a move away from passive consumption towards active engagement, with technology enabling new forms of interaction and co-creation in music.
The rapid advancement of AI in music creation is outpacing the current music rights infrastructure, necessitating new models for IP management.
Blockchain technology could offer a robust solution for the challenges posed by AI in music creation, providing mechanisms for consent, control, transparency, and fair compensation of IP use.
Introduction
The music industry has been introduced to a wide array of AI tools and platforms that span the entire creative process, from initial composition to the final stages of distribution and marketing. These tools and platforms vary in complexity, from basic recommendation systems and assistive AI tools to advanced generative AI models.
In a study evaluating the influence of AI on the music creation process for two leading European collecting societies, GEMA/SACEM, it was found that 63% of participants anticipate AI will be predominantly used in composition, lyric writing, and various other creative tasks. Following that, 58% believe AI will play a significant role in recording, editing, mixing, and mastering, while 55% see its potential in creating promotional content.
The more advanced Generative [1] AI models have rapidly risen to prominence as the principal point of interest in the music industry. These models are seen as a double-edged sword – an opportunity for increased creativity and productivity versus a severe threat to copyright infringement and the dilution of music. While this prevailing discussion rightfully exists, the real potential of generative AI is found elsewhere – accessibility. Generative AI is set to transform how consumers experience music, moving beyond consumption to a medium of expression and entertainment accessible to all.
[1] Generative AI music refers to AI models capable of generating music autonomously or semi-autonomously.
The new creator model – The consumer creator
Dennis Kooker, President, Global Digital Business, Sony Music Entertainment:
“We’re at the beginning stages of a very significant transformation in how people want to experience and consume music.”
The music industry is known to evolve and transform in response to the predominant medium of consumption, from vinyl records and CDs to the dominance of streaming. Each transition radically changes the way we experience and interact with music. With Generative AI, the industry is presented with a completely new wave of disruption beyond consumption – it is catalysing a new creator model driven by consumers.
Consumerisation of music creation
Reviewing the slogans of some of the top generative AI tools and applications provides insight into their overarching vision:
“A future where anyone can make great music. No instrument needed, just imagination. From your mind to music.” – Suno4
“Create original songs in seconds, even if you‘ve never made music before.” – Boomy2
“Add a soundtrack to your latest Instagram post with ease.” – Meta/MusicGen3
They place a significantly greater focus on consumers over existing musicians – targeting those who have not previously had the opportunity to express themselves through music. Generative AI opens up new forms of creativity, enabling anyone to produce music – ushering in a new era for the traditionally passive consumer.
This new period of consumer behaviour can be compared to the similar advancements seen in other creative fields – such as photography and film. Unlike traditional photography, which requires specialised knowledge and equipment, smartphones enable anyone to create, share, and consume high-quality photos. This ease of use altered the perception of photography from a hobby or profession to a prevalent part of daily life. Social media platforms, notably Instagram and TikTok, developed around this trend, resulting in photos and videos becoming the primary modes of sharing experiences, expressing emotions, and telling stories. AI-generated music caters to different facets of media consumption and creation than traditional music. Just as YouTube, TikTok, and Netflix serve various functions for video, AI-generated music and traditional music will embody various types of entertainment and cater to diverse needs, preferences, and contexts within music.
The participatory culture of Gen Z
Gen Z is adept at leveraging technology and social media to connect, organise, and interact in new ways that previous generations could not. This participatory role has led to the growth of ‘participatory culture’. This culture sharply diverges from traditional consumer culture by positioning individuals not just as consumers but as active and engaged participants. Consequently, the rise of Gen Z’s “participatory culture” will further enhance the shift towards casual, consumer-driven music creation.
“A participatory culture lowers the barriers to artistic and civic engagement, encouraging every voice to contribute, share, and learn from one another.” - Henry Jenkins
With participatory culture and AI rising, the artist-fan relationship will evolve through casual music creation. Traditional fan engagement is passive, with fans acting as observers, listeners, or consumers, contributing little to no input. As fan engagement becomes active, a more interactive two-way relationship between creator and consumer is driven by co-creation, remixing, and music (re)creation.
Imagine fans bringing their favourite music into their gaming experiences. Depending on the game’s nature and gameplay—whether action-packed, sombre, or leisurely—the music’s characteristics (such as tone, pitch, tempo, and ambience) adjust accordingly. Fans could then share these custom tracks on community platforms, enabling others to explore new remixes of the artist’s work and vote on the best versions for each game scenario or setting. Through this process, artists can effectively produce game soundtracks fueled by community involvement. Fans contributing to this process are rewarded, cultivating a stronger bond with the artist.
The role of the creator economy
The creator economy is thriving and expected to approach half a trillion dollars by 2027 – encompassing over 50 million creators and serving hundreds of millions of consumers. Casual music creation has yet to succeed in the creator economy despite its success. This stems from the fact that music creation within social media is limited, mainly confined to TikTok duets and Snapchat’s Sounds feature.
Reflecting on the evolution of photography, the creator economy tools and platforms have opened up a broad array of entertainment around it. For instance, creating a video on TikTok or Instagram empowers users to modify aspects such as saturation and exposure or apply various filters. Yet, editing options are absent when incorporating audio into images or videos. Music IP management has yet to develop alongside the dynamics of the creator economy to support these features. Imagine if these social media platforms expanded their toolkit to include music editing features, enabling users to slow down tracks, isolate or remove vocals, swap out instruments, and introduce background sounds that complement the photo or video.
Just as TikTok introduced new forms of short-form video content, AI opens avenues for exploring innovations in music that haven’t been possible through traditional means. It encourages experimentation, which could lead to the discovery of new musical genres, forms of creativity, and modes of expression. The area of consumption for music is likely to adapt and segment according to these different forms of casual music creation. Streaming services like Spotify are already overwhelmed with abundant content; consumer AI creations won’t find a home here. Just as there are spaces for both Netflix’s curated content and TikTok’s user-generated videos, new environments will emerge tailored for AI-generated music use cases.
The caveats – Why generative AI needs blockchain
The advancements in AI significantly outpace the industry’s ability to adopt it. The current rights model must be built to handle this trend’s widespread licensing needs. The shift towards Generative AI and mass consumer-driven music creation stretches the boundaries of our current understanding of music rights but also necessitates the introduction of entirely new categories of rights. The industry needs a flexible and dynamic model that supports innovation, protects IP, and allows mass licensing on a global scale.
The music rights infrastructure is not built for our digital age
Music rights were not designed for AI; even streaming was beyond their initial scope. The music industry’s rights model isn’t built for today’s digital landscape. This is clearly illustrated by streaming’s lack of a distinct right – combining mechanical and public performance rights instead.
Traditionally, music rights were pretty straightforward – if you bought a CD, the musicians got paid for that copy. But with streaming, things got more complicated. A single stream of a song is treated as both a copy of the song (like buying a CD) and as a performance, meaning different types of payments must be made. This double nature has led to burdensome workarounds to ensure artists get paid without making it too hard to share music online. Even though we call it streaming, legally, it’s like juggling two balls (copy and performance) with one hand.
Introducing AI-generated music doesn’t just add complexity; it multiplies it. Think of the original scenario, where consuming music through streaming involved juggling two types of rights: the right to copy the music and the right to perform it publicly. Now, with AI stepping in to casually create music, we’re adding more “balls” to our juggling act, making it more like juggling six balls at once. When an AI creates music, determining the holder of this right becomes a puzzle. Is it the developer of the AI, the user who prompted the creation, a combination of the two, or a new legal entity altogether? How the payment flows as a result is yet another aspect to consider.
One emerging challenge will be the rise of content that exists only temporarily. It’s plausible that a significant portion of future music content will be temporal, existing briefly before disappearing—much like social media “stories” or videos that vanish after 24 hours. Despite its temporary nature, this content also has the potential to generate revenue, underscoring the need for a rights framework that ensures creators are adequately compensated.
Another challenge is defining a new synthetic personality right. If an AI is designed to emulate a specific artist’s style, how does this infringe on that artist’s rights? For example, if an AI tool copied specific melodies or lyrics, that would likely constitute copyright infringement. However, it is difficult to identify such specific examples of copying, with well-built AI tools generally designed to copy the more general sound and feel of music, in part to avoid allegations of copyright infringement. This consideration extends beyond just an artist’s style and applies to companies producing synthesisers. What are the implications when AI models blend the distinct sounds of a Moog or Juno synthesiser?
Patent rights could add yet another dimension, with the potential for the algorithms and technologies powering AI music generation to be patented. This introduces another set of considerations regarding who holds these patents and how IP ownership of the training data affects the dissemination and monetisation of the music produced.
What is needed for generative AI to thrive?
To accommodate the age of Generative AI, it’s essential to modernise the music rights system and IP infrastructure to align with today’s digital environment. Rightsholders should be equipped with mechanisms for consent, control, transparency, and fair compensation (TuneCore). Similarly, consumer creators need straightforward access to licensing systems that are as simple to use as creating AI content.
With the surge of creators across the Internet, establishing a system that offers clear and effective IP and rights management is crucial. This system should facilitate digital ownership, licensing, distribution, and traceability of IP, ensuring the responsible and ethical utilisation of generative AI in music creation.
Consent and Permission for IP Use
Rights holders should have a say in how their work is used in AI-driven music creation. A flexible permission framework allows creators to specify their terms for using IP in AI remixing, co-creation, and sampling. This approach respects the intentions and rights of the original creators while fostering innovation and collaboration in music creation.
Transparency of IP Use
Many algorithms powering Generative AI in music are trained on vast datasets of copyrighted IP. Transparency regarding which data is used for training these models and how it is used is important for the integrity of the creative process and the protection of creators’ rights.
The Provenance of IP Use
Beyond tracking the origin of IP for AI model training, it’s crucial to establish a system that meticulously tracks the usage of AI-generated music. This system should identify the origins of samples, vocals, and instruments used in AI-generated music and record how and by whom these elements are employed. This detailed provenance supports accurate recognition of all contributors in the music creation process, from original creators to those who generate derivative works through AI.
Fair Compensation of IP Use
Original creators and rightsholders must be acknowledged and compensated fairly for the exploitation of IP. Developing compensation models that reflect the contribution of each party in AI music creation is vital. These models should account for the new ways music is created and consumed, ensuring that all participants in the creative process are rewarded for their contributions.
Empowering AI While Protecting IP with Blockchain
Blockchain technology solves several challenges posed by integrating Generative AI into the music industry – supporting the shift to a consumer-creator era. It represents the infrastructure needed to implement a system that provides efficient and transparent IP ownership, licensing, distribution, and provenance mechanisms.
The transparent, decentralised, and immutable nature of blockchain, combined with tokenisation, a process of converting rights to an asset into a digital token, enables rightsholders to establish undisputed digital ownership of their intellectual property. While most of blockchain focuses on payments, its fundamental purpose is verifying authenticity. This tokenisation process, usually in the form of a non-fungible token or NFT, means that all information surrounding the IP is digitally preserved about who created what, who owns what, and where it is used.
Blockchain-Enabled Consent and Permission
Blockchain’s inherent properties enable rights holders to programme authorisation and usage rights into their tokenised IP. This allows rights holders to securely and transparently specify the terms of use for their IP. Smart contracts on top of blockchain infrastructure can automate the enforcement of these permissions, ensuring that the use of copyrighted materials aligns with the creator’s terms. This system respects the rights of original creators and streamlines the process for AI-driven music creation, making it more efficient and compliant.
Verifiable Record of IP Usage
Blockchain’s transparent and immutable nature makes it an ideal technology for tracking the use of copyrighted IP in AI model training. Every transaction on a blockchain is recorded in a permanent and easy-to-verify way, enabling a transparent and auditable trail of how data is used. Furthermore, by integrating tokenised IP into blockchain-based applications, the origin and usage of AI-generated music can be tracked, creating a transparent and unchangeable record of each piece of content’s journey. This not only aids in establishing the authenticity of AI-generated music but ensures that all contributors, from original creators to those continuously extending derivatives through AI, are accurately recognised and compensated.
Automating Fair Compensation Through Smart Contracts
Smart contracts perform certain automated processes like transferring ownership, managing royalty distribution and compensation, or licensing copyrights for AI model training. These smart contracts are based on predefined rules, ensuring fair treatment to all parties involved in the music creation process. These contracts can be programmed to account for the nuances of AI-generated music, where traditional compensation models may not adequately reflect the contributions of all parties. This approach ensures that creators, whether providing original content or utilising AI to create derivative works, are rewarded appropriately for their contributions.
Blockchain offers a foundation for modernising the music rights model and IP infrastructure to suit the needs of the digital and AI-driven era. Its characteristics address the core requirements of consent, control, transparency, and fair compensation.
Parting Thoughts
With AI opening up a new age of creativity, the industry is witnessing the birth of a new creator model, where music creation is not just the domain of traditional musicians but also a new form of entertainment accessible to the everyday consumer. This shift towards consumer-driven music creation mirrors the development of casual consumer photography, suggesting a future where producing high-quality music will be just as easy as taking a picture. While this may sound like the opposite of creativity, Generative AI will unlock a new standard for music, allowing anyone to use it as an expressive medium rather than strictly a medium of consumption.
The future role of Generative AI in music creation demands a rethinking of the industry’s foundational structures and principles. By addressing these needs—ranging from provenance tracking and transparent use of training data to fair compensation —the music industry can embrace the transformative potential of AI. Blockchain technology highlights the necessity for innovation in rights management to protect rightsholders while fostering the growth of the consumer creator. By supporting consent, control, transparency, and fair compensation, blockchain provides the infrastructure to ensure the industry can thrive in this new digital landscape.
References:
Goldmedia. (2024). AI and Music Report. gema.de. https://www.gema.de/documents/d/guest/gema-sacem-goldmedia-ai-and-music-pdfBelieve/TuneCore. (2023). AI in Music Report. https://www.tunecore.com/wp-content/uploads/sites/12/2023/07/TuneCore-AI-Study-Report-1.pdfMulligan, M. (2023, May 25). AI will unlock creation rather than consumption. MIDiA Research. https://www.midiaresearch.com/blog/ai-will-unlock-creation-rather-than-consumptionMulligan, M. (2023b, August 31). AI, Music Rights, and known unknowns. MIDiA Research. https://www.midiaresearch.com/blog/ai-music-rights-and-known-unknowns