More about our work

Copyright law is critical in protecting artists’ creations, from their lyrics, to their musical tracks, to even the concepts of their albums. The concept of copyright and reproduction guidelines extends as far back as the 1500s with the invention of the printing press (Aziz). Formal adoption of copyright law has existed since the 1700s, with two landmark laws laying the foundation for copyright law around the world: the Statute of Anne passed in Britain in 1710, and the Berne Convention for the Protection of Literary and Artistic Works in 1886 (Geller). The former was the first instance of public governmental involvement in copyright protection, and the latter was a historic international agreement on how creators could control the usage of their works. Copyright law has changed with the moving landscape of music, as music publishing increased, and new technology allowed for greater reproduction of music recordings.

While copyright law has historically been sufficient, artificial intelligence (AI) fundamentally challenges the notion of an “author,” and who or what qualifies as one. Copyright law thus far assumes that authors are human, but this no longer holds true in music. AI themselves are producing creative works, leading to two key questions (Drott):

1. Who owns an AI-created work? Is it the programmer who made the AI, the artist who used the AI, or should the AI itself hold some rights? If it’s the case that the AI holds ownership, then traditional copyright law is inapplicable, and a whole new legal framework must be established.

2. Should AI be allowed to train on copyrighted material? Because AI algorithms are trained on millions of data points, this training data becomes the basis of the algorithm’s behavior. Oftentimes, these data points include copyrighted material (Drott). As such, the AI algorithms may output results that are similar to the copyrighted material from which it learned.

Current copyright law is unclear on how to handle either situation. AI is redefining the sphere of music, and with it, the copyright law that protects it, leaving much uncertainty over licensing, copyright, and creative rights.

Our goal is to ensure that artists and creators are protected, but to responsibly use AI to foster a new wave of music and advancement, revolutionizing the industry through a new wave of music composition and performance previously unimaginable.

Works Cited:

Aziz, Atif. "Artificial intelligence produced original work: A new approach to copyright protection and ownership." European Journal of Artificial Intelligence and Machine Learning 2.2 (2023): 9-16.

Drott, Eric. "Copyright, compensation, and commons in the music AI industry." Creative Industries Journal 14.2 (2021): 190-207.

Geller, Paul Edward. "Copyright history and the future: what's culture got to do with it." J. Copyright Soc'y USA 47 (2000): 209.

The state can influence the development of AI technology in music through three primary roles: as a builder, regulator, and enabler.

BUILDER

As a builder, the state would primarily allocate funding and grants. By providing financial support for existing research institutions and private corporations to research and develop these algorithms, the state can spur innovation, allowing for more innovative solutions and faster growth. In doing so, the state grants freedom to the most qualified organizations to research and develop as they see fit, and fulfills its unique duty in facilitating the pursuit of scientific advancement (Bush).

However, the state as a builder politicizes the process, potentially introducing political hurdles and corruption. Because the state determines where the funding goes, certain political interests and connections can influence the decision-making processes. As such, the state may end up unfairly prioritizing organizations that best help the agenda of politicians, or favor those with the greatest lobbying efforts, instead of doing the greatest for public interest.

REGULATOR

As a regulator, the state would regulate how developers can train AI music algorithms, and how people can use AI for music. For example, the state could mandate developers not use any copyrighted music for training, or ban certain uses of AI such as mimicking known artists.

By regulating training and utilization of AI, the state can protect the interests of artists and their property rights. As a first step, this would require greater solidification of how AI fits into copyright law, as to know how best to regulate AI. Further, the state could monitor the control of companies with AI algorithms, breaking up any that become a monopoly, thus allowing for greater competition and benefits to consumers (Wu).

As with much governmental regulation, the fear is that inappropriately restrictive guidelines or breaking up companies could stifle innovation by preventing AI from being developed or used a certain way, or curtail people’s freedom of expression and creativity. In addition, regulation requires follow through and monitoring, auditing of companies, and a general resource expenditure to ensure people obey the regulations.

ENABLER

As an enabler, the state would clearly detail and enforce copyright and intellectual property, determining who owns these AI created works. In doing so, the state provides a standard framework for artists to rely on to protect their intellectual property. By knowing their rights, artists protect their dignity and confidence, knowing exactly how they can use AI while still maintaining legal ownership, and knowing exactly how others can use AI on their copyrighted material.

The difficulty in being an enabler for intellectual property is that the AI landscape is constantly evolving and unknown, with complex AI use cases involving multiple parties. There are currently many legal ambiguities for what ownership looks like with AI created works, and these ambiguities will only continue as AI evolves in ways we do not expect. Enforcing copyright is also a great difficulty because of the ease in creation and distribution online of AI music. Finally, because both artists and AI take inspiration from others, a challenge in enforcing intellectual property is determining the difference between a breakthrough work, versus an incremental copy of someone else’s work (Beuchamp).

OUR STANCE

We recognize the importance the state has in shaping what AI in music looks like. That’s why the state must be primarily a builder. While regulation and enabling are important, the state's role as a builder is exceptionally important because of the constant advancement of AI music technology – it’s not even close to its full potential; we’re just scratching the surface of what AI technology is capable of. While regulation and enabling are crucial as well, these only matter insofar as there is a clear sense of what the industry looks like. For either to matter, the state being a builder comes as a prerequisite.

By allocating funding to us, the State can guarantee that it’s driving innovation in not just AI music, but AI technology as a whole. AI is taking over every facet of life – music needs to keep up. Investing in us will be a catalyst for future revolutionizing discoveries, and transformative advancements that change what we fundamentally think of as music. The State as a builder demonstrates a commitment to innovation and economic growth, and thus bettering people’s quality of life. We can innovate, and the State can help.

Works Cited:

Beauchamp, Christopher. "Who invented the telephone? Lawyers, patents, and the judgments of history." Technology and Culture 51.4 (2010): 854-878.

Bush, Vannevar. Science, the endless frontier. Princeton University Press, 2020.

Wu, Tim. The curse of bigness: how corporate giants came to rule the world. Atlantic Books, 2020.