Artificial Intelligence remains a relatively novel aspect of technology, rapidly advancing at an exponential pace, with seemingly limitless potential. However, this progress brings forth a unique set of ethical dilemmas, often centered around issues such as copyright and the reliability of AI systems. As AI continues to evolve and permeate the music industry, these concerns become increasingly significant. Today’s AI can convincingly replicate any voice, applying it to entirely different lyrics or spoken words. Although this represents a remarkable technological breakthrough, a voice is an intensely personal and distinctive attribute inherently linked to the individual. Artists either possess an innate talent or invest years honing their vocal skills. The use of AI also sparks many legal debates regarding ownership and “fair use;” can an individual copyright their voice? Or does that exceed the existing legal frameworks for tangible forms of authorship? In reflecting upon “Thinking with Bad Bunny,” it is crucial to examine how AI’s role in music might encroach upon the rights of artists whose voices are replicated and utilized without their consent. Considering Bad Bunny’s brand, which is deeply rooted in authenticity and his free-spirited approach, it is essential to consider how AI interpretations might influence his identity and brand image. Bad Bunny has also built a brand on authenticity and the idea that “he does whatever he wants;” how do interpretations of AI affect him and his branding specifically? Through the analysis of several literary sources and potential interviews, I intend to answer the question, how does AI impact the socially accepted norms and ideas of ownership, especially as it pertains to artistic expression and the use of one’s voice?
Monica Huerta –The Unintended: Photography, Property, and the Aesthetics of Racial Capitalism
There are many articles and texts related to the complex interplay of legal frameworks, ethical considerations, and technological advancements that influence the shaping, ownership, and use of photographic images. A primary theme across all sources is the tension between the photographer’s creative labor and the subject’s right to their image. Monica Huerta’s The Unintended: Photography, Property, and the Aesthetics of Racial Capitalism (2023) examines the intersections between photography, property, and racial capitalism. The book delves into how photography has been historically utilized in shaping perceptions of property and race, highlighting the unintended consequences of these visual representations. Huerta explores the cultural and social impacts of photography, analyzing how images can both reinforce and challenge notions of ownership and racial identity in the context of capitalism. Through her analysis, she seeks to uncover the complexities and contradictions inherent in visual culture and its role in social structures.
McGowan’s article, “The AI Artist Who Used Bad Bunny’s Voice—and Shot to Fame,” adds a contemporary dimension to the debate by exploring the use of AI in music production. It profiles Mauricio Bustos, a Chilean artist using AI to create music under the moniker FlowGPT. Bustos employs AI voice cloning to generate songs featuring the vocals of popular artists like Bad Bunny and Justin Bieber, sparking controversy and a cease-and-desist from Bad Bunny. Despite the backlash, FlowGPT’s music gained millions of streams. The use of AI voice cloning to replicate the styles and voices of famous artists, while raising concerns about copyright infringement, also reveals the transformative potential of technology in challenging traditional industry structures and potentially democratizing music creation. This, however, also introduces new ethical quandaries regarding consent and authorship.
Policy Surrounding AI Use
What is “fair use” and why does it hold so much relevance in the debates about AI and copyright? Fredrickson Law Firm provides a legal update titled, “Protecting Voice in the Age of AI,” which delves into these topics and what kind of legal action can be taken. Ultimately, not much can be done as a voice is not protected under copyright law because it is not seen as a tangible aspect of an individual’s Name, Image, and Likeness (NIL). There is no federal law that protects a person’s voice, so it is up to the states to decide whether the misappropriation of a voice was “distinctive,” “widely known,” and for monetary or commercial gain. In an article from The Washington Post titled “AI’s future could hinge on one thorny legal question,” they debate the idea of “fair use” and whether it is legal to use New York Times articles that are copyrighted to train OpenAI technology without consent. Does this qualify as “intellectual property?” Again, it is blurry whether or not true legal action will come to pass in regards to AI. The article states, “Broadly speaking, copyright law distinguishes between ripping off someone else’s work verbatim — which is generally illegal — and “remixing” or putting it to a new, creative use.” This idea of remixing is especially prevalent in musical spaces because it is a common practice. Where AI diverges is its lack of the original content creator’s consent when utilizing their sound. There is also a distinction between borrowing a clip of someone’s voice versus training technology to clone that voice and have it say/sing whatever you like without that individual knowing.
Huerta’s historical analysis in her article and book provides a crucial foundation for understanding the current debate surrounding image rights. Her work effectively traces the evolution of legal frameworks designed to protect individual rights, yet simultaneously reveals their limitations. McGowan’s article on AI music production introduces a new layer of complexity, showing how rapidly evolving technology complicates existing notions of authorship and copyright. This highlights the need for a proactive, rather than reactive, approach to regulating AI-generated content, ensuring equitable treatment of both artists and subjects. In terms of the legal framework for “fair use” in regard to AI, it is still yet to be seen if these lawsuits will be successful and what precedents will be set in terms of AI usage, especially in creative spaces.
Singify AI Music Song Generator
“Music Creation Has Never Been Easier”
As a way to illustrate how simple it is for anyone, including someone with no advanced AI experience, I will be utilizing AI technology to impersonate Bad Bunny’s voice and make him sing in English. Bad Bunny has outwardly spoken his opinion on AI and it was not positive, yet his voice still remains a prompt on the AI song-generating website, Singify. The model has already utilized machine learning based on Bad Bunny’s music and likely interviews and has almost 30k uses and 135 likes, indicating that thousands of copies of his voice have been made despite his distaste for the practice.
The example song even has him singing in English. What makes Bad Bunny’s voice unique also makes it easy to clone and that is where a lot of the discomfort around AI voice technology is coming from. This enhances conceptual understanding of how easy it is to do this and how well it can come out with little to no AI experience. However, it is important to recognize the imperfections in this technology as well. For my first track, I inserted a clip of Sabrina Carpenter’s song, Juno, which is expertly edited and sung in perfect key. The AI responded very well to this polished track and was able to insert Benito’s voice in a relatively seamless way. However, Singify had a much harder time understanding more raw tracks and vocals. For my second example, I wanted to utilize an English translation of one of Bad Bunny’s most famous songs, Tití Me Preguntó, which was relatively easy to find online. Upon entering these vocals, I noticed many more discrepancies and areas where it is no longer convincingly Benito’s voice. It is worth recognizing that these are not the most high-end AI
tools, but it appears that regardless, the inserted tracks need to be the highest quality possible in order to produce quality and convincing tracks.
Juno AI Cover
Titi Me Pregunto AI Cover
Tití Me Preguntó (English Translation)
The Importance of Voice
The voice is, in my opinion, one of the most uniquely personal aspects of expression, especially when it is used as a musical instrument. The tone and language a person uses identify them and give queues to that person’s background. The voice can also be seen as a political act as the words a person uses and the way they phrase them can reveal more than just the surface level of information they are trying to convey. When Benito says, “Yo hago lo que me da la gana,” or “I do what I want,” all of these aspects of voice matter. This phrase isolated in text can be interpreted as standoff-ish or selfish, but when you hear Bad Bunny speak it aloud it is understood that it comes from a place of creative freedom with his music and how he conducts himself. He doesn’t let any external forces dictate the way he navigates the world or his sound and it is one of the things that makes his brand so unique. When he does fall into c
ertain stereotypes or “normal” modes of behaving, he simply responds in the same way, “Yo hago lo que me da la gana.” I appreciate that his version of authenticity doesn’t restrict him to being different or outside the social boundaries all the time; if he enjoys something that is mainstream, he allows himself to enjoy it and brush off the criticism that it is “basic” to do so. Notions of voice and its importance to an individual play a significant role for Bad Bunny specifically since his voice is so unique and easily recognizable. While this makes it potentially more challenging for another human to replicate, his signature style makes it easy for AI to clone. Bad Bunny’s adverse reaction to his voice being taken and used in an AI-generated song by Mauricio Bustos (McGowan, 2024) seems to stem from an infringement upon his authenticity. I interpret Bad Bunny’s emotional reaction to a sense of a loss of control over something very personal to him; if he can’t control how his voice is being used and the messages it sends, what can he do? The use of his voice in this context almost feels like a parody of him rather than who he really is.
Conclusion
The rapid advancement of artificial intelligence in the music industry raises critical ethical concerns, particularly regarding copyright and the ownership of one’s voice. As we have seen through the examination of literary sources and the practical implications of AI technology, the uniqueness of an artist’s voice is not merely a technical attribute but an integral part of their identity and self-expression. Bad Bunny’s reaction to AI-generated music that uses his voice without consent underscores the profound emotional and personal impact this technology can have on artists, challenging notions of authenticity and control over their own sound. By exploring the background of AI policy and its implications for artistic ownership, we can better understand the delicate balance between innovation and the rights of creators in this evolving landscape. As AI continues to shape the way music is produced and consumed, it is essential to critically assess how these changes affect not just the artists themselves, but also the broader cultural narratives surrounding ownership and artistic expression in our society. My multi-media project shows just how simple the use of AI technology is to create convincing voice clones and essentially force artists to say whatever I want. By having Bad Bunny sing in English, I am changing an integral part of his sound and the politics of his music. AI impacts ideas of ownership and authenticity and also leads me to question everything I see in the media, as it is not as challenging as one may think to completely clone someone’s likeness.
Bibliography
Editorial, F. (2023, October 10). The unintended: Photography, property, and the aesthetics of racial capitalism. https://faculti.net/the-unintended-photography-property-and-the-aesthetics-of-racial-capitalism/
Huerta, Monica, The Unintended: Photography, Property, and the Aesthetics of Racial Capitalism. New York: New York University Press, 2023.
Huerta, Monica. “Monica Huerta ’06 Highlights Legal Consequences of Photography in Latest Book.” Princeton Alumni Weekly, August 21, 2023. https://paw.princeton.edu/article/monica-huerta-06-highlights-legal-consequences-photography-latest-book.
Huerta, Monica. “What’s Mine: Involuntary Expressions, Modern Personality, and the Right to Privacy.” The Journal of Nineteenth-Century Americanists, vol. 4, no. 2, 2016, pp. 359–89.
Mack, Zachary. “If Property Is Theft, Then What Is a Photograph?” Lamb
da Literary Review, December 6, 2023. https://lambdaliteraryreview.org/2023/12/if-property-is-theft-then-what-is-a-photograph/.
McGowan, Charis. “The AI Artist Who Used Bad Bunny’s Voice—and Shot to Fame.” Rest of World, 9 July 2024, https://restofworld.org/2024/ai-song-flowgpt-bad-bunny/.
YouTube. The Unintended: Photography, Property, and the Aesthetics of Racial Capitalism – Monica Huerta. YouTube. https://www.youtube.com/watch?v=-ko5s3mswbk
Protecting voice in the age of AI. Fredrikson & Byron P.A. (n.d.). https://www.fredlaw.com/alert-protecting-voice-in-the-age-of-ai
YouTube. (n.d.-c). YouTube. https://www.youtube.com/watch?v=XIg9fNaw0rE&t=7s
YouTube. Tití Me Preguntó by Bad Bunny (ENGLISH TRANSLATION). YouTube. https://www.youtube.com/watch?v=ZRB00z8XJD4
1000+ ai voice models: Fineshare Singify. 1000+ AI Voice Models | Fineshare Singify. (n.d). https://singify.fineshare.com/covers/ai-bad-bunny
