Intersection Of Copyright And AI - Assessing The Legality Of Sampling AI-Generated Music In India
Author: LawSphereX
Category: Technology, Cyber & Privacy
Published:
Updated:
INTERSECTION OF COPYRIGHT AND AI: ASSESSING THE
LEGALITY OF SAMPLING AI-GENERATED MUSIC IN INDIA
Jitasha Bahl, O.P. Jindal Global University
ABSTRACT
The advent of AI in music production raises complex challenges for copyright law, particularly in India, where traditional frameworks inadequately address the nuances of AI-generated works. This paper examines whether sampling AI-generated music should be deemed illegal, arguing against such a stance due to the absence of originality and human authorship in AI outputs. After evaluating three potential approaches— defining authorship, requiring transparency, and placing AI-generated works in the public domain—the paper advocates for the latter as the most viable solution. This framework fosters innovation, avoids legal ambiguities, and ensures a balanced integration of AI and human creativity in music production.
INTRODUCTION
The advent of AI has marked a significant technological advancement but has resulted in several implications and uncertainties in intellectual property law. The main question that this blog seeks to address is whether sampling of artificially generated music in India should be legal. The broader question is whether traditional copyright enforcement mechanisms are adequate to address these new forms of infringement by AI generated content. It is evident that the current copyright laws are insufficient due to lack of adaptation to new technologies but the real question is how to cope with this rapid progress in AI. Due to the restrictions in applying traditional intellectual property frameworks, there is a unique and complex challenge for lawmakers to revise laws to keep up with these technological advancements and remain relevant in the digital age1.
In light of this, three potential frameworks are explored for overcoming these complications. However, as these are not without their own limitations, this paper suggests the most viable option for moving forward. SAMPLING OF AI GENERATED MUSIC AI-generated music is transforming the creative landscape of music production. Traditionally, artists have relied on their skills, personal experiences, and emotions to compose songs. However, with the advent of AI, this process has been streamlined, as AI can suggest chords, melodies, and even lyrics. While the effect on the originality of music is debated, AI opens up opportunities for musicians to explore new creative possibilities. For beginners, AI offers a way to experiment with and create music without needing significant prior knowledge or experience.
Professionals can use AI to reduce the time spent on routine tasks, enabling them to concentrate more on their artistic vision. This technology could democratize music creation, making it accessible to anyone with access to a computer. Sampling a musical work without obtaining prior permission from the original copyright holder constitutes a copyright infringement. However, to prove such an infringement, the copyright holder must not only establish their ownership rights but also demonstrate that the infringer has copied a substantial portion of the original work. The technique of sampling AI-generated music represents a unique intersection of technology and creativity. It challenges traditional notions of copyright and ownership, as AI-generated music lacks the human input that has historically been central to
1 Soni, T., 2024. Impact of AI on IPR Framework. Available at SSRN 4831898.
copyright protection2. THE LIMITS OF TRADITIONAL COPYRIGHT MECHANISMS The Indian copyright legislation is not equipped to handle work created by AI. Despite the glaring insufficiencies of the current copyright laws, the Union Ministry of Commerce and Industry expressed no interest in amending copyright laws and stated that India’s existing Intellectual Property Rights (IPR) regime is well-equipped to protect AI-generated works and does not require a separate category of rights3. The Indian Legislature has revealed that it has no intentions of amending the laws or adapting them to the needs of the digital world. Nonhuman entities are becoming more integrated into human society, however, it remains undeniable that they are not considered natural persons. As a result, human-made laws may struggle to adapt to this evolving reality.
Redefining copyright authorship to include nonhuman creators would disrupt the existing copyright framework. Below are the main challenges that could arise if copyright protection were granted to AI for its creative outputs4.
1) Defying the Traditional Theories of Intellectual Property Rights Protection-
Granting copyright protection to AI Generated works contradicts the theories of intellectual property law which are the core principles for justifying protection to creative works. The Incentive Theory posits that recognizing or rewarding creators for their work provides an incentive for further innovation, enriching society by motivating individuals to contribute intellectually. This theory, however, is not applicable to AI systems as they do not require incentives like human creators, and as non-sentient beings, they cannot derive any personal or professional gain from the recognition of their output. As AI does not benefit from ownership or rewards, it invalidates the rationale for granting it copyright. Even without copyright protection, AI would continue to function in the same way, making copyright unnecessary.
Further, the Labour Theory which justifies rewarding human physical and mental effort is also absent in AI-generated work. AI performs tasks autonomously without any conscious effort, making it challenging to justify granting copyright to the AI itself. The Personality Theory, 2 Music Sampling And The Defence Of Doctrine Of Fair Use - Copyright - Intellectual Property - India, https://www. mondaq. com/india/copyright/1158044/music-sampling-and-the-defence-of-doctrine-of-fair-use. 3 Existing IPR regime well-equipped to protect AI generated works, no need to create separate category of rights, https://pib. gov. in/pib. gov. in/Pressreleaseshare. aspx? PRID=2004715. 4 Indian Journal of Law and Legal Research, Analyzing The Scope Of Copyright Protection For AI Generated Works, IJLLR (2023), https://www. ijllr. com/post/analyzing-the-scope-of-copyright-protection-for-ai-generatedworks.
which asserts that an individual's will and personality are expressed through the creation of intellectual works is also lacking since AI lacks its own personality or will and is entirely dependent on the programmer’s instructions, data inputs, and training5.
2) Fails to Meet the Criteria for Creativity under the Indian Law-
Copyright law in India, as established by the Supreme Court in Eastern Book Company vs. D. B. Modak6 (2008), requires that a work demonstrate skill, labour, and judgment to qualify for protection. AI-generated works, however, fail to meet these criteria. The creative skill required to produce a work originates from the programmer, not the AI. AI is simply executing tasks based on the data it has been fed. AI does not "labour" in the human sense; it only processes data and performs tasks autonomously, without any personal effort. It also lacks the capacity to make independent creative judgments because its output is predetermined by the programming and data provided by humans, meaning any creative judgment comes from the programmer, not the AI.
In light of these factors, AI-generated works cannot qualify for copyright protection, as they do not meet the standards of human creativity required under Indian law7.
3) Challenge
to the Human-Centred Design of Intellectual Property (IP)- Firstly, under Indian law, copyright protection for original literary, dramatic, musical, and artistic works lasts for the life of the author plus 60 years after the year of their death. However, if AI were to be recognized as the author of a copyrighted work, the rationale for the duration of copyright protection becomes moot as there would be no point at which the copyright would expire. This could theoretically lead to indefinite copyright protection, disrupting the balance that copyright law seeks to maintain between rewarding creators and eventually allowing works to enter the public domain8. Secondly, AI's ability to generate vast amounts of content with minimal effort raises questions about the economic fairness of granting it copyright protection. Unlike human creators, who 5 Ahrenstein, Andrew.
"AI Generated Art and the Gap in Copyright Law. " Am. U. Intell. Prop. Brief 15 (2023): 23.
6 Eastern Book Company & Ors vs D.B. Modak & Anr, AIR 2008 SC 809.
7 Stuti Puri & Seema Gupta, AI and Copyright
Navigating Legal Frontiers in the Age of Artificial Intelligence, 4 JUS CORPUS L. J. 26 (December 2023 - February 2024). 8 Martin Clancy, Law: You Can Call Me Hal: AI and Music IP, in Artificial Intelligence and Music Ecosystem (2022). invest significant time, labour, and creativity into their work, AI can produce content rapidly and without exhaustion. Granting AI copyright protection could flood the market with AIgenerated works, potentially devaluing human-created content. This would create an imbalance in the creative economy, where AI-generated works dominate, and human creators find it increasingly difficult to compete. Thirdly, since AI lacks legal personhood, it cannot initiate legal proceedings to enforce its copyright in cases of infringement.
Legal actions require a legal entity to assert rights in a court of law, and without personhood, AI is incapable of performing this fundamental function. Moreover, AI may unintentionally generate content that infringes on existing copyrights, given that it often relies on large datasets to create new works. Holding the AI liable is impossible because it cannot understand legal concepts such as liability and responsibility would likely fall on the developers or users of the AI, even if they had no direct involvement in the specific content generated9. POTENTIAL PATHWAYS One possible approach is to establish a framework that clearly defines authorship in the context of AI-generated music. This would involve recognizing both the human programmer and the AI as co-authors of the work, reflecting the contributions of both parties in the creation process10.
While defining authorship by recognizing both the human programmer and the AI as co-authors may seem like a reasonable approach, there are significant drawbacks and challenges to implementing this model. Introducing a dual authorship system for AI-generated music could lead to substantial legal uncertainty, particularly when disputes arise over the degree of contribution made by the human versus the AI. The process of assigning ownership percentages between the human programmer and the AI is fraught with complications. The percentage allocation would depend on various factors, such as the complexity of the AI’s algorithm and the amount of human input involved. However, determining these percentages in a consistent and fair manner across different cases would be difficult. A second approach focuses on transparency and open data.
Under this model, AI developers would be required to disclose the training data and algorithms used to create their AI music
9 KHAN, M. and PRAKASH, D.D., 2024. " THE INTERSECTION OF AI AND COPYRIGHT IN MUSIC
INDUSTRY. IJRAR-International Journal of Research and Analytical Reviews (IJRAR), 11(2), pp.362-436.
10 P. Karthikeyan & B. V. Tamizhkavi, AI in Incorporeal Property Works
Helper or Squelcher, 1 LAW ESSENTIALS J. 2 (January-March 2021). tools. This framework would emphasize ethical practices, promote innovation, and address legal concerns about potential plagiarism and originality. However, AI systems are often proprietary, and the data and algorithms used in their development are considered valuable. Forcing developers to disclose this information could undermine their competitive advantage and deter innovation. Many AI companies invest heavily in developing unique datasets and algorithms, and requiring full transparency could lead to the reverse effect, where businesses are reluctant to share their innovations for fear of losing intellectual property protection. The third approach involves establishing a system where AI-generated music is placed in the public domain after a certain period.
This would promote remix culture and encourage further experimentation and innovation in music creation11. One of the primary benefits of this approach is that it effectively overcomes the limitations of authorship and originality inherent in traditional copyright frameworks, which are built around the concept of human authorship. By placing AI-generated works in the public domain, there is no need to assign authorship to a non-human entity, thus removing the dilemma of whether the AI or its developer should be credited as the author. Moreover, the public domain model mitigates the challenges related to copyright duration by ensuring that AI-generated works are placed in the public domain, there is a prevention of indefinite copyright protections for non-human entities, thus promoting ongoing cultural and creative development.
The public domain serves as a reservoir of creative resources, allowing future creators to draw upon previous works to inspire new ideas and innovations, thereby enriching the cultural landscape. SUGGESTIONS Of the three options for establishing a new legal framework—defining authorship, requiring open data and transparency, and public domain access—the most viable solution is to place AIgenerated music in the public domain after a certain period. While both Defining Authorship and Open Data and Transparency Requirements offer potential pathways to regulating AIgenerated music, each presents significant challenges that make them difficult to implement effectively. Meanwhile, the option of Public Domain Access and Innovation effectively addresses the traditional limitations of copyright law by removing barriers related to authorship, duration, enforcement, and ethical use.
Allowing AI-generated music to enter the 11 Seeking Authorship in AI-Crafted Content, A Fruitless Endeavour? Exploring the Realm of AI-generated Music | IPLINK ASIA, https://www. iplink-asia. com/article-detail. php? id=1178 (last visited Oct 15, 2024). public domain after a defined period would foster creativity, promote remix culture, and avoid the complexities of assigning ownership to non-human entities. While transparency mechanisms could complement this framework by ensuring ethical AI development, the emphasis on public domain access aligns with both the needs of creators and the evolving landscape of AI-generated content. The case for placing AI-generated music into the public domain is not without criticisms, but these criticisms can be neutralized when we consider the broader benefits.
By placing AI-generated music in the public domain, India can maintain the incentive structures that reward human creativity, while avoiding the legal and philosophical issues that arise from trying to assign ownership to non-human entities. Public domain access encourages cross-pollination of ideas, enabling greater innovation through collaboration, and prevents the market from being flooded with too many similar AI-generated compositions. Additionally, this approach helps avoid the legal controversies surrounding authorship and copyright infringement, as placing AI-generated music in the public domain eliminates the need for complex legal determinations about who owns the rights to an AI-generated work12. This, in turn, reduces the likelihood of litigation and promotes a more open and accessible music industry.
Allowing AI-generated music to enter the public domain would not only preserve the distinctive significance of human artists but also foster an environment where human creativity and AI technology can work together harmoniously. As artists increasingly rely on emerging technologies to push the boundaries of the artistic experience, placing AIgenerated music in the public domain ensures that these collaborations remain open and accessible. This approach allows for a confluence of human creativity and technological sophistication, enriching both human culture and artistic traditions. Rather than viewing AI as a competitor to human creativity, placing AI-generated works in the public domain positions AI as a valuable tool for enhancing human expression.
It also encourages an ongoing dialogue between art and technology, ensuring that the future of music production remains innovative and inclusive. THE WAY FORWARD Scientific progress is exponential, particularly in the field of AI, meaning that any changes made to existing copyright laws to accommodate AI-generated works would likely become obsolete in a short period. AI is not only advancing in the artistic domain but across various 12 Nordås, Hildegunn Kyvik. "Copyright and trade in the digital music industry. " In Handbook of Innovation and Intellectual Property Rights, pp. 177-190. Edward Elgar Publishing, 2024. fields, including science, medicine, and technology, which implies that any legal framework created today to address AI-generated content would soon be inadequate.
As AI technology evolves, its capacity to generate creative works, mimic human behaviour, and operate autonomously will increase, leading to even more complex legal challenges. A small adjustment to current laws, such as granting AI copyright or creating a new category for AIgenerated works, would quickly need further revision as AI becomes more sophisticated13. This constant need for legal updates could result in instability and confusion within the legal system, as legislators struggle to keep pace with AI developments. Public Domain Access, on the other hand, provides a more sustainable solution in the face of rapid technological progress. By placing AI-generated works in the public domain, the legal system avoids the need for frequent revisions and updates. It creates a stable, long-term framework where AI-generated content can be freely used and built upon by the public.
This approach allows the legal system to maintain flexibility without being overwhelmed by the pace of AI innovation14.
CONCLUSION
The rapid development of AI-generated music presents significant challenges for copyright law, particularly in India, where traditional enforcement mechanisms are inadequate for addressing the complexities of AI-driven creativity. Given the inadequacies of traditional copyright enforcement mechanisms, sampling AI-generated music should not be illegal in India. AI-generated works lack the originality and human authorship required for copyright protection, and banning their sampling would stifle creativity and innovation. Instead, India should adopt a more flexible legal framework that allows for the free use of AI-generated music, promoting artistic expression and technological advancement. This paper has explored the complex issues surrounding the applicability of copyright protection to AI-generated musical works.
An analysis of statutory law indicates that extending these protections to AI is challenging, as copyright laws are inherently human-centric and designed to protect human creators. Expanding copyright to include AI-generated works could result in overprotection, which might paradoxically undermine the incentives necessary for fostering innovation and creativity. The public domain has been identified as a potential solution and is widely favoured
13 Supra note 5.
14 Kop, Mauritz. "AI & intellectual property
Towards an articulated public domain." Tex. Intell. Prop. LJ 28 (2019): 297. by academics. Stronger legal protection for AI-generated works should only be considered if AI is explicitly granted some form of legal personhood which seems very unlikely15. 15 Akanksha Majumdar, Facing the Music: The Future of Copyright Law and Artificial Intelligence in Music Industry, (2023), https://papers.ssrn.com/abstract=4754032.