At the 2025 Paris AI Summit, VPOTUS Vance declared to world leaders that “excessive regulation” harms the AI industry and will not be tolerated by the USA. His position contrasts with another VP, that of the world’s largest music company (Universal), who denounced AI’s “wholesale hijacking of the intellectual property of the entire creative community.” Indeed, as Suchir Balaji showed, the “fair use” doctrine cannot reasonably apply to the “training” of AI, whether in the form of text, images, or music, since the resulting synthetic products are designed to compete commercially with human-made creations. Law has arguably not kept pace with new technologies, including music AI, which flagrantly violates the spirit of copyright. How are ethnomusicologists to respond to AI in ways consistent with our values? Currently, the US, China, and Europe are the main centers of AI innovation, and of these the EU most explicitly protects privacy and AI safety (e.g. GDPR, EU AI Act). The US is also one of the only major countries that is not a signatory to major international agreements for safe AI development. Since SEM is a US-based organization, its members must consider the impact these US policies will ultimately have on music ownership and music creation worldwide. Based on a decolonial approach to IP in the context of international law, this presentation will identify established ethnomusicological values, then outline the legal arguments (and counterarguments) for regulating AI to protect musicians, promote cultural survival, and even ensure the future of human personal identity.

Court decisions regarding Madonna and Kraftwerk might yet shape the AI music-copyright debate. Of late, copyright's legal balance between corporation and individual has shifted to occlude the author, even as fair-use provisions have been extended, in US and EU jurisdictions, to protect Silicon Valley corporations' practices of harvesting, caching, indexing, and thumbnailing public data online. This paper thus asks how we might better understand "data mining" as a fair-use exception in light of a landmark decision by the US Ninth Circuit, on minimal sampling, and a pending decision by the Court of Justice of the European Union (CJEU), on pastiche. Together, the two cases summon urgent questions regarding legal constructions of authorship, artistic agency, and fair use.

While the decision in VMG Salsoul v. Ciccone (2016) promises a renewal of the de minimis doctrine pertaining to sampling of minimal length or extent, the pending CJEU decision, on appeal, in Pelham GmbH, et al. v. Ralf Hütter and Florian Schneider-Esleben (2023) promises to rehabilitate the fair-use exception for pastiche. These cases might thereby clarify music-AI copyright: to what extent do training models "borrow" fragments of recorded music under copyright? And, to what extent do AI audio generators reassemble them in their outputs? While Sebastian Stober (2024) has argued that notions of borrowing and pastiche are technically incommensurate with AI audio generators, I argue that—whether or not jurists fully comprehend them—such decisions might yet deliver a legally binding understanding that will shape the direction of music copyright, creativity, and industry to come.

This paper explores what it might mean to "decolonise" AI in the music industry, drawing from case studies in the UAE and Egypt. AI in the music industry is used in music recommendation, to detect copyright infringement, and in music production, as only a few examples. Scholars such as Georgina Born have critiqued the Western centric nature of AI, with most dominant AI systems in the music industry developed by and for Western pop music. Using the Arabic music industry based out of the UAE and Anghami, a Beirut-based music streaming service, as a case studies, this paper demonstrates how the Arabic music industry operationalises AI in ways that both challenge and uphold these relations of power within the region and beyond. Rather than understanding AI as an isolated and coherent object that can be decolonised, it questions whether AI, as a socio-technical assemblage always embedded in its larger cultural and historical contexts, might be better approached through the notion of the colonised coloniser. Developed by historian Eve Troutt Powell, the concept of the colonised coloniser expresses a position of duality in which one both liberates and dominates simultaneously, where the very drive toward liberation is justified through a logic of domination. Most broadly, this paper questions the possibility of decolonising AI, showing its limitations even when starting from the epistemologies of listeners and musicians in the global South. This paper is based on one year of fieldwork in the UAE and Egypt between 2022 and 2023.

This presentation will explore the phenomenon of Artificial Intelligence (AI)-created popular vocal music in the years 2023 through 2025. Through examining both music creation and consumption, I will examine how AI softwares perform humanity-- the results of which purposely blur the lines between AI and human. I also delve into complications that arise when AI performs race and gender. The rise and fall of AI-rapper FN Meka will be used as a case study through which I consider the ethics of an AI software performing Black masculinity with goals of profit and commodification, and other case studies will be used as well. I will also provide potential future developments in the relationship between AI and music, and its impacts on human musicians and listeners.