AI-Music Cloning Debate

Introduction

Artificial intelligence is enabling unprecedented transformations across many domains due to rapid progress, including within music sector. Among the most sophisticated recent AI developments is voice cloning with perfect mimicry, presenting the music industry with novel creative, ethical, and legal difficulties to address. Replicating a vocalist’s intonations and timbers with absolute accuracy introduces modern challenges for original composition while additionally accentuating longstanding unsolved issues regarding attribution of rights and ownership over artistic works and personas. Furthermore, mimicking musicians motivates examination of where attribution ends and derivation veers into inauthentic imitation, blurring lines of copyright law and personality rights. In this blog the author has tried to understand the impact and implications of music cloning in the music industry while using Artificial Intelligence. This blog also focuses on protecting the copyright, personality rights and moral rights of the original creator while emphasizing upon the required changes to address these digital and modern complexities.

Understanding AI Music Cloning

Cloning music with artificial intelligence requires deep-learning techniques to recreate a singer’s unique vocal qualities from extensive datasets of their vocal recorded performances. The advanced systems are capable of simulated nuanced aspects such as vocal tone, cadence, style and emotive undertone which could render impersonations indistinguishable from the original performances. These potential uses also includes the creation of new compositions or collaborations having the vocals of deceased legends, without involving the original vocals. AI music cloning is not only limited to using vocals and styles of famous or deceased singers, but it also encompasses the vocal tone of distinguished or known personalities. One of the prominent examples, is the viral meme on social media, demonstrating the AI-generated music featuring the vocals of Indian Prime Minister Shri Narendra Modi. Although, such usages are digitally appreciated as a part of meme or parody, but at the same time alarms the unethical use of AI-music cloning technology.

The usage of AI-music cloning technologies has sparked a storm of debates and controversies relating to whether the usage of such a technology is infringing or not. Some views that these technologies have the ability to skilfully emulate the human vocals which will be helpful to produce new compositions, while the others consider this as a copyright infringement and violations to the moral rights[1] of the original creators or vocalists. Recently, the AI replication of the vocals of late singer Mohammed Rafi,[2] highlighted the ethical challenges of artistic integrity and legacy of such deceased artists.

Copyright Implications of AI-Generated Music

The primary function of copyright law is undoubtedly to protect creative works that are clearly established in a physical, tangible form. It awards the creators of such works exclusive rights[3] to reproduce, to distribute, and to derive further works from their original conception. The copyright being considered as bundle of rights, it confers various ownership rights in a song. The lyrics of a song is considered as literary work and the lyricist have the right over the lyrics of the song.[4] The composer gets the right in relation to the musical work,[5] the producer being responsible for the sound recording[6] is considered author in respect of that sound recording,[7] and the singer who is considered as the performer[8] gets the performance rights. However, the AI-music cloning songs have controversial notions relating to authorship[9] and ownership[10] rights over the AI-generated songs and music. The AI-generated songs face complex issues relating to its transformative use and fair use, as these songs being the replication with pixel-level accuracy, fails to fall under the exceptions to copyrights law. The major challenges include:

1. Originality: The primary necessity of copyright law is ORIGINALITY.[11] But while considering AI-generated music it raises concerns of “originality”, as the consideration of AI generated work as original is still a debated issue, and any developer or user cannot claim it to be an original creation.
2. Expression: We know that the copyright subsists in expression of ideas and not in idea themselves.[12] Here while we talk about the expression of works duplicated by AI, the determination of authorship rights over such an expression stays uncertain. As to whom authorship rights will be determined, the developer, user, or the original artist is still debatable and dubious.
3. Derivative works: The term derivative works means a work based upon any existing work. Hence, the AI-generated music comes under the purview of derivative works. Although, for derivative works to come into existence, permission and authorization of the original creator or artist is required, as failing to do so, the work will potentially infringe the original creator’s right. As most of the AI-music cloning technologies fails to have the consent and prior permission of the original artist (whom vocals are being taken), makes it very uncertain and questionable to determine an AI-generated music as a derivative work. Moreover, recently famous singer Arijit Singh filed a case against an AI-music cloning company which used the vocals and style without his permission and consent.

Personality and Moral Rights

An artist’s voice embodies their identity, protected by personality rights in some jurisdictions. These protections safeguard individual’s name, likeness, and vocal tone from unauthorized commercial exploitation.[13] In India, though personality privileges are not explicitly codified, courts have acknowledged them through cases such as Arijit Singh v. Codible Ventures.[14] This landmark case was a pivotal judgment where the Bombay High Court limited the utilization of Arijit Singh’s artificially cloned voice, accentuating the rational and lawful necessity to protect performers.

Similarly, moral rights guarantee creators protection from harm to reputation by preventing distortion or mutilation of their works. Section 57[15] asserts that no one may distort works to exploit the reputation of original creator. Simulating an artist’s unique style without consent is against the principles of moral rights as enshrined in the copyright laws. AI-generated contents such as songs or mimics of an artist’s voice without their consent, contravenes the moral rights such as right to integrity and paternity, especially when used in context of harming the reputation, image and value of an artist or creator.

Global Legal Frameworks and Case Studies

Globally, regulations surrounding AI music cloning varies significantly between jurisdictions:

1. United States: The US recently introduced the “No AI Fraud Act” to safeguard and protect creators and their work from any kind of unauthorized AI-generated work. In the famous case of Midler v. Ford Motor Company,[16] the US copyright legislation shows limitation in protecting the voices of artists. Although the petitioner Bette Midler, was successfully granted protection for her right to publicity, to avoid any unauthorized replication by using her distinctive vocals to promote Ford vehicles. This case showed up that US Copyrights law alone cannot protect the voices of artists, but it safeguards the artists or creators from any such unauthorized AI-generation abilities.
2. European Union: The EU have their own AI act, which helps the creators and individuals from any kind of AI-generative work, which is against the original owner’s consent and permission. The EU’s initiative helps to protect and balance the rights of original creators and the modern day innovation, while adhering with the law.
3. India: The Indian copyright law protects the performer’s ownership right for 50 years post-performance.[17] Although, the current legislation doesn’t have any specific provision safeguarding the voices of artist to secure it from AI-unauthorised replications and cloning. In the famous case of Experience Hendrix LLC v. Purple Haze Records Ltd,[18] the court held that performances are protected under the copyright legislations, but was silent upon the protection of vocals of artists.

Ethical Concerns and Industry Response

In this AI world, the music industry faces various complex challenges such as AI music cloning, which clashes with the ethical dilemmas and rights of an artist. The AI music replication is mostly termed as derivative work, but the most important concern is with regard to consent, as using an artist’s vocal without their consent or permission undermines the artist’s right and autonomy. In one of the recent examples, the voice of deceased Anthony Bourdain was used for creating a documentary[19] without any clear permission and authorization, which provoked various objections to the work. Such an uncertainty creates various risks and challenges with respect to transparency and accountability, as the listeners are mostly unaware and deceived into believing the AI-generated work or music as original work of such artist whose vocals are being used. Although, if developers start putting disclosures and disclaimers, as a measure to ensure transparency, this could help to address these concerns.

The other most important issue pertaining to the unauthorised usage of AI generated work, is related to the artistic legacy and personality rights that they have, especially for the deceased artists. As the AI innovation to create new and unique compositions and music tracks by making clone or replication of artist’s voice sparks nuanced debates relating to the protection of integrity rights of the artists. As, the Copyrights legislation is uncertain and silent while protecting the voice and vocal tones of artists, the heirs and managing companies of such artists often face this complex tussle between safeguarding the legacy, dignity and integrity of artists and at the same time the tension of commercializing the artist’s vocals. And in this era of unethical usage of AI, the protection to artist’s integrity while safeguarding their right over their vocals and voice tunes are crucial to protect their legacy, and this is more vulnerable when this debate moves around preserving the rights of a deceased legend. Now to tackle such issues, the prominent music labels of the industries such as T-series, Sony, etc., have started engaging with the developers to establish licensing agreements, which aims to safeguard the IP rights of the developers as well the moral rights of an artists, protecting their original works from any unauthorised replication or cloning.[20]

Recommendations for a Balanced Framework

While we discussed various challenges and complexities while safeguarding the artist’s rights and their original works, and making a way for the future prospects of AI generated works, it is clear that the legislation needs comprehensive legal reforms. The Copyrights legislation requires provisions relating to protection of vocal replications of artists to protect the artist’s vocals from any unauthorized AI usage. Also, the IP laws need to define authorship and ownership rights in AI-generated works, so that liabilities can be ensured in cases of violations. The policy makers also need to determine the rights over the artist’s voice posthumously. The provisions should be implemented in such a way that it safeguards the artistic legacies and rights while also fostering the AI innovations.

One of the challenges with respect to safeguarding the artist’s right in this framework is transparency, as the origin of AI-generated works are still uncertain and ambiguous. The legal frameworks should be drafted in such a way that it certifies the ethical development and application of AI-generated work in the music industry, which will be helpful to determine accountability. Further, royalty should also be given to the original artists whose voices are cloned by AI, to commercialize the newly composed tracks with the help of AI-music cloning technologies. An overall balance is required to oversight both the technical and socioeconomical perspectives of these AI-generated songs.

Lastly, it is established that the moral rights including the integrity and paternity rights along with the reputation and legacy of artists are violated by such unethical and unauthorised use by AI-music cloning technologies. So, industry collaborations with developers and public awareness will be helpful to address the issues of these emerging technologies. Developers, performers and policy makers, all together can bring a positive change and solution while protecting both the rights of the original artists as well as acknowledging the new and unique creations of AI-generated songs via ethical pathway. As progress demands prudence – not blanket acceptance nor rejection, but a thoughtful navigation considering multiple valid viewpoints would be crucial to find solutions relating to AI-music cloning challenges.

Conclusion

The AI-music cloning technology is undoubtedly innovative, but it comes with various challenges and complexities. Although the technology can be used for various exciting advancements, but its misuse and misapplication such as unethical and unauthorised usage endangers the rights of an artist over their original work. Although, judiciary in various of its landmark judgments had safeguarded the rights of artists but right to vocal reputation, is something which is still not much explored. The judiciary have given protection to the artists unique reputation in terms of personality rights, similarly there is a requirement to protect and safeguard the vocal and voice notes of singers and artists, as their vocals are the main crucial thing which requires an ultimate protection. In this age of artificial intelligence, some stringent regulations relating to AI music cloning is must to safeguard the rights of artists along with the fostering the innovations of such AI technologies, while respecting the contributions of the original creators.