A deep-dive into the Principles for Music Creation with AI.
Part 1 of this story focused on the journey to AI For Music. This post will look at the principles themselves and offer some additional context and explanation from me.
Disclaimer: The views and opinions expressed in this story are solely my own and do not necessarily reflect those of my employer, colleagues, or any affiliated organizations. This post is intended to share a personal perspective from my experience working on this project.
The Principles for Music Creation with AI were co-developed by Roland and Universal Music Group. While they draw inspiration from vanguard efforts such as the Human Artistry Campaign, these principles have been authored with all music creators in mind. We wanted the language to be as plain as possible, and for the core messages to be immediately understandable by anyone.
As we worked through the process of building the principles, the litmus test I would use for myself was: “can I imagine these principles being discussed and understood by friends in a music store, between sets at weekend band gigs, on social media”, and so on. And while there is nuance in some of the language (which this article will explore), we’ve received very few questions about interpretation since the launch of AI For Music in March of 2024 (and I don’t think this the result of apathy).
Let’s dive in…
We believe music is central to humanity. Enjoying and creating music is integral to our health, well-being, and happiness. Music provides a human connection to each other and the world around us and is deeply personal.
expansion: This is an opening declaration that making and enjoying music holds irreplaceable value as part of the personal and tribal human experience. If you’ve read this far, it’s unlikely you need to be convinced of this!
We believe humanity and music are inseparable, and music with true meaning cannot exist without the direct creative contributions of people.
expansion: Of all the principles, this is the one I felt might generate the most discussion, specifically concerning the interpretation of true meaning and the need for human contribution. While I explore this topic more deeply in “Can AI create music with emotion” part 1 and part 2, briefly, I believe that meaning comes from human stories, and it’s the presence of this meaning in music that inspires deep and lasting emotional connections.
From acoustic to electronic instruments, we believe that technology has long supported human artistic expression, and applied sustainably, AI will amplify human creativity.
expansion: This statement simply affirms the synergistic relationship between music and technology throughout history, and looks with optimism toward the potential of AI to enhance human creativity when developed and applied responsibly and sustainably. Indeed, speaking for Roland and humbly on behalf of UMG, both companies are deeply invested in creative AI development and are excited for the future.
We believe that human-created works must be respected and protected, and they uniquely merit copyright protections. The use of copyrighted works – and use of the music artists’ name/image/likeness/voice by others – must be authorized in advance of the use and music artists should be given recognition for their contribution.
expansion: This principle speaks to the heart of what is being argued in courts around the world right now. It affirms that under copyright law, creators and assigned rights holders must decide how their work can and cannot be used, including for the training of AI models. Furthermore, rights of publicity protect an artist’s name, image, likeness, and voice, and these protections should extend to the training of AI models.
We believe transparency is essential to responsible and trustworthy use of AI. In order to establish and maintain trust across all stakeholders, from fans to artists, AI-powered creative platforms and policies must prioritize record keeping and disclosure while shining a light on those who create.
expansion: At the time of this writing, many (if not most) AI platforms are black boxes, where training data sources are either not disclosed to the public or are discussed under the cover of heavy legal jargon, often buried deep within a website’s navigation. To build and maintain trust with the public (artists and fans), a paradigm shift toward openness and transparency from AI providers, specifically regarding model training data, must take place.
We believe the perspectives of music artists, songwriters, and other creators must be sought after and respected as AI-powered creative experiences are imagined, strategized, planned, and implemented.
expansion: At Roland, we talk about the risks of innovating in a vacuum with AI, specifically, where and how AI might be welcomed by artists into their creative workflows, and just as importantly, the areas to be avoided. Needless to say, there is a growing volume of research on this topic, but the spirit of this principle is about engaging with artists directly and often. Since launching AI For Music in 2024, we’ve already seen community engagement activities from several supporter brands (including Roland and UMG, of course!), but more is needed, and those efforts must be sustained as views and sentiments continue to evolve along with the technology.
We are proud to help bring music to life. Together, through the musical instruments we produce, the artists and songwriters we represent, and the music education programs we support, we inspire and enable creativity.
expansion: This final principle serves more as a closing statement; a confirmation that supporters of AI For Music wholeheartedly embrace the incredible, life-changing opportunity to partner with artists and inspire the world to express themselves creatively through music.
Please use the comments section below to ask questions or let me know what else you’d like me to write about on this incredibly important topic!