Last Date for Submission - 30th August

MUSIC AND INTELLECTUAL PROPERTY RIGHTS IN THE ERA OF AI AND STREAMING

By S Jayavardhini
Student, The National University of Advanced Legal Studies, India.

Email id: sjayavardhini2167@nuals.ac.in

The relationship between intellectual property (IP) and music has always been dynamic, influenced by changing legal frameworks, creative processes, and technological advancements. Traditional ideas of authorship, ownership, and originality have been called into question by the emergence of artificial intelligence, which has caused a paradigm shift in the production and dissemination of music in recent years. Questions concerning who is considered an author, how current intellectual property laws apply to AI-generated content, and how equitable compensation can be guaranteed in a music industry that is becoming more digital and algorithm-driven are raised as artists, producers, and platforms navigate this quickly evolving terrain. With an emphasis on the urgent need for legal reform to preserve a balance between innovation and creator rights in the digital age, this paper explores the challenges surrounding copyright protection, streaming royalties, and cross-jurisdictional disparities.[1]

The copyright act 1957 defines “originality” as the first and foremost requirement for copyright to subsist in a work. The work must originate from the skill and labour of the creator and must not be taken from any other source for it to be deemed as original. The work must also be in a tangible form[2]. The copyright extends only to the original expression of ideas and not to the ideas itself according to the fundamental rule laid down in Article 9(2) of the Trade Related Intellectual Property Rights (TRIPs)[3] and Article 2 of WCT, 1996[4] which the Indian regime follows. Musical work includes any work comprising of music and includes graphical notation (representation of music through visual symbols outside the realm of traditional music notation) of such work but does not include words or actions intended to be sung, spoken or performed with music according to section 2(p) of musical work act[5]. Sound recordings refer to recordings of sounds from which sounds may be produced irrespective of the medium or method used to generate such recordings. The category of works that qualify for copyright protection is a closed, however the definitions within each category of work is intended to be extensive and inclusive accommodating evolving forms of innovativeness and expression[6]. Section 22 of the Musical Works act provides the duration of copyright in musical works published within the lifetime of the author until sixty years from the beginning of the calendar year next following the year in which the author dies. If it is a case of joint authorship, the term shall be counted at or immediately before the date of the death of the author who dies last[7]. Copyright however doesn’t persist if the sounds recording produced with respect to any literary, dramatic or musical work if copyright itself has been infringed in making the sound recording.

In January 2025, the U.S. copyright office released part 2 of its report “Copyright and Artificial Intelligence” which showcases how copyright law applies to AI generated content. This report reaffirms that human creativity plays a key role in copyright protection in the United States[8]. Mere AI generated output lacks meaningful human creativity and the necessary authorship required for protection under the copyright act. This aligns with the existing frameworks, case laws and administrative decisions including recent copyright office rulings denying registration for purely AI generated works. (find case laws). The report also addresses hybrid authorship scenarios where human contribution must be considerable, evident and independently copyrightable extending beyond mere trivial prompts or trivial contributions. Although the report provides examples, ambiguities and doubts still exists in defining “sufficient human creativity” such as would small changes to AI- generated text or images such as adding a personal style or touch still qualify as sufficient human creativity or must there be substantial transformative effort? This report thus creates uncertainty especially for businesses and developers who rely on AI tools for their business and it may lead to increased dependence on contractual agreements for their business such as licensing terms that define how AI must be used or distributed[9].

Another important requirement for AI generated work to be granted copyright protection is authorship. Author is defined as “he to whom anything owes its origin; originator; maker; one who completes a work of science or literature[10].” In the case of Thaler v. Perlmutter (2023)[11], the court concluded that copyright protection is only reserved for works with substantial human creation and innovativeness. In this case, Stephen Thaler made an effort to apply for copyright protection for his work which was generated by an AI algorithm called “creativity machine”. The court conducted an in-depth analysis of the definitions of authorship in legal sources such as the constitution and held that authorship requires human origin. As of March 2024, the copyright office issued its latest guidelines after a TikTok user with the username “Ghostwriter977” produced a hit song- ‘Heart on my sleeve’, using music of superstars Drake and the Weeknd, stating that AI work without any human involvement or intervention cannot be copyrighted[12]. Platforms such as TikTok showcase the beginning of a new revolution in AI generated art where millions of users and listeners can be reached opening the doors for other AI systems to create art based on songs.

However, the court has failed to address certain grey areas such as when human authors prompt AI tools to generate music. For example- lets consider a situation in which a person writes a song using lyrical elements of hit songwriter Justin Bieber, while another person prompts AI to generate a song in the style of Justin Bieber based on the existing database of Bieber’s lyrics. In the former case, it is a clear threat of copyright infringement under human authorship clause but in the latter scenario, under the current legal guidelines, artists such as Justin Bieber cannot sue AI developer, user or the system for copyright infringement. Currently the US copyright office Statement of policy expresses the sentiment that if a human prompts an AI system to create a work in which the AI has complete control over the expressive elements of its output, such work doesn’t fall under the category of human authorship and thus is not copyrightable[13]. It can be observed that the Copyright Office’s human authorship clause does not make room for human-prompted AI generated music and is only broad enough to regulate human beings. However, the rapid increase of AI generated music in platforms such as Spotify, YouTube and TikTok means that human creativity is being intruded by AI systems everyday and makes it necessary to designate a party to be held responsible for copyright infringements in such niche situations.

In 2020, AI generated versions of Jay-Z’s voice were used to create popular songs and the deepfake versions were so convincing that it raised the questions about unauthorized use of an artist’s voice without permission posing significant challenges to copyright and personality rights law[14]. This also raised significant questions about whether rights belonged to AI developers, users or AI itself. In the case, Experience Hendrix LLC v. Purple Haze Records Ltd[15], it was held that a singer’s voice could be protected through the performance even though the artist’s voice in isolation is not generally protected. In the landmark case of R.G. Anand v. Delux Films[16], it was held that copyright infringement is determined based on whether an average listener after hearing both the works gets an unquestionable impression that the subsequent work is a copy of the original work. In case this test fails, it is crucial to examine whether it could fall under the defense of ‘Fair dealing’. The case The Civic Chandran v. Ammini Amma[17], laid down the components for fair dealing such as quality of the matter taken, the purpose for which it is used, and the likelihood of commonalities between the works, placing importance on the transformative nature of the work even when an existing person’s music heavily influences their creations. Yet current definitions of ownership and originality are insufficient for AI-generated music leaving ambiguities and neglecting to address some important areas requiring thorough legal updates to balance innovation with the rights of the creators.

Traditional rights management has been radically reshaped by streaming services and digital licensing, which has presented a number of new difficulties for creators, rights holders, and the industry as a whole such as complexity, lack of transparency, royalty distribution, digital piracy etc. There exists a lot of confusion in the music industry due to the complex nature of streaming deals, the method of licensing differing between record industry and music publishers, the method in which services are licensed being different from country to country, unpredictability of the revenue sharing arrangements between most streaming deals etc. Copyright laws usually define ‘presumed’ of ‘default’ owners of new works but default owners can usually transfer ownership for return of money through assignment or licensing agreements. Copyrights can also be co-owned which is quite popular in the music industry as collaboration is common in songwriting. In such a case, a licensee must take permission from each and every stakeholder to make use of the work[18].

Modern music streaming platforms have transformed content licensing practices, shifting from traditional linear broadcasting to an on-demand consumption model. Streaming is no longer territory based. Global licensing strategies are now necessary for streaming, which is no longer based on geographic boundaries. Traditional licensing methods and valuations need to be re-examined in light of this paradigm shift. Streaming services now use exclusive content as a primary differentiator which has led to growing prices for exclusive licensing agreements and development of techniques for striking a balance between non-exclusive and exclusive content. This makes it necessary rethink conventional windowing techniques to optimize content value in the age of streaming and to keep up with the latest developments that will influence streaming and content licensing in the future[19].

It has also come to light that streaming favours already privileged or well-known artists giving them the upper hand in terms of payment and compensation. The amount of pay that a right-holder receives depends on the number of views and the listens that they receive or what the market share is. This method of compensation also known as “pro rata” compensation tends to favour popular artists while the newcomers are left struggling. The streaming platform keeps 30% of fees from its subscribers and ultimately split the remaining fees among the right-holders of the recordings of the song (performers, record label and producers). The share that each of the many parties involved in a song's production ultimately receives is extremely small. In contrast, independent artists typically lose market share because they lack the same resources as record labels, even though they may not be required to pay a share. As streaming has become more popular, concerns regarding rights management have grown. Understanding how licensing agreements operate is crucial for any business or person involved in the music industry, as streaming is changing the way royalties are paid out[20].

Different jurisdictions are responding to the IP implications of AI-generated music in different ways with a mix of continuous discussion, gradual reforms, and legal conservatism with some regions experimenting with different strategies or making the current frameworks more understandable. As discussed earlier, the report published by U.S copyright office makes it clear that solely AI-generated content is not copyrightable and a person doesn’t get the title of an ‘author’ by only giving prompts to AI. Copyright status is also evaluated in a case-by-case basis with clear proof of human contribution (such as writing lyrics, layering, post-production work etc.) and there is no shortcut or blanket protection directly available[21]. For example- in the case of Randy Travis and AI music where Randy Travis used AI to restore his voice, copyright was approved due to direct human involvement in production and arrangement process, thus making it clear that the stronger the human role, the stronger the protection[22]. The legal frameworks in the European Union is similar to that of the United States. According to the European Union Intellectual Property Office and the recent regulations of the Court of Justice of the European Union, evidence of the distinct level of human involvedness and intellectual creation is necessary for copyright protection. Through the reCreating Europe project funded by European Union's Horizon 2020 research and innovation program, the EU is exploring measures to ensure transparency in AI generated work and promoting innovation while respecting copyright norms[23]. Certain tools such as AIVA have been formally acknowledged as tools for music creation, significantly departing from traditional norms on authorship and copyright laws[24].

In the United Kingdom, under the Copyright, Designs and Patents Act 1988, the UK grants authorship to a person who made arrangements for producing output even though the output is computer-generated and may lack human authorship. But since the use of AI is increasing and the definition of ‘person who made arrangements’ is becoming more and more equivocal, this framework of the act is being called into question. The UK government's recent plans to permit AI systems to train on creative works from enormous databases without the original creators' consent have also caused a great deal of resentment among many artists in the country. A group of UK musicians, many of whom were well-known, protested the free use of their work by releasing a silent album. It has been viewed as a symbolic act to criticize the proposed changes, highlighting what artists perceive to be a loss of their copyright protections for technology, reflecting underlying tensions over how to strike a balance between creative rights and technological advancement[25]. In Australia as well, copyright protection is currently restricted to work created by humans defined as independent intellectual effort. In Australia AI has no legal rights or personhood, there is also no obligation to reveal that a song is AI generated and there is no legal precedent to determine who owns AI generated music. This is a legal copyrightable area unless new standards and laws are established[26].

India's copyright system, like that of the majority of common law nations, is based on originality and human authorship. RAGHAV artificial intelligence was the first ever artificial intelligence tool that was recognised by the Indian copyright office as an AI painting app as the co-author with ownership rights given to Ankit Sahni who authorized the painting ‘Suryast’[27]. However, subsequently the Indian Copyright Office issued a withdrawal notice to Ankit Sahni as it did not fall into the definition of ‘author’ in relation to artistic work under sections 2(d)(iii) and (iv) of the Copyright Act. When Mr. Sahni filed for protection in the jurisdictions of USA and Canada, the AI tool was granted copyright protection by the Canadian Intellectual Rights Office (CIPO) but not in the USA. The grant of copyright protection in one jurisdiction but rejection in other leads to international inconsistencies and uncertainties[28].

To ensure fair compensation and protection for music creation in this digital age the introduction of several key reforms is necessary many of which are already under debate or consideration. A new bill was brought before the US congress which intends to increase royalties which are paid to artists for their music streams. A new standard or minimum wage was set for musicians making sure that every artist was paid at least “one cent per stream” thereby protecting all artists and musicians and not just the top and established ones. The aim of this bill is to protect artists who have been directly affected by the music industry's lack of regulation. Congresswoman Tlaib points out that existence of inequality in the music industry leaving many artists behind who are struggling to make ends meet[29].

In 2023, Congresswoman Deborah Ross introduced Protect Working musicians Act[30] which will give small independent musicians and producers the ability to collectively bargain for just compensation with streaming services and generative artificial intelligence developers. Under current laws, small and independent musicians do not have an ability to bargain for market value rates for the usage of their music on international streaming services like Apple Music and Spotify. They have virtually no control over AI companies that frequently scrape and use their music without their knowledge or consent, and they are compelled to accept whatever terms these platforms offer. Without being constrained by antitrust laws, this law enables independent musicians to organize and engage in collective negotiations with major streaming services and AI developers. The Artist Rights Alliance (ARA) and the American Association of Independent Music (A2IM) support the Protect Working Musicians Act[31].

The Music modernization act 2018 is an example of an act that has had to change and adjust to the emerging technology. The act aims to address legal loopholes and challenges put forth by digital streaming and transform the existing copyright regime. Prior to the MMA licensing regime there was exploitation to a great extent in the music industry where digital music providers could make huge profits without paying just and fair royalties to the creators. The MMA however brought in a ‘blanket mechanical licensing’ where instead of obtaining a separate license for each work, the blanket license will permit providers to broadcast work available in a set by paying the appropriate fees. Global expansion and improvement of such systems would guarantee correct payments, lessen administrative burdens, and assist in resolving disputes pertaining to unclaimed royalties[32]. n order to reduce confusion and lost revenue in cross-border digital music use, international cooperation is required to create unified, transparent rights databases and licensing procedures. Because music is now consumed and distributed worldwide, it is crucial to harmonize copyright and royalty laws across jurisdictions to avoid legal snags and guarantee that musicians are compensated and protected wherever their music is streamed.

In conclusion, the shortcomings of the current intellectual property laws have become increasingly evident as artificial intelligence continues to influence the music industry. The ambiguous realm of AI-generated works is difficult for the current legal frameworks, which are based on human creativity, to handle, leading to international inconsistencies and a lack of clarity on ownership and liability. Additionally, the emergence of digital streaming platforms has exposed glaring disparities in artist compensation, which frequently marginalize independent creators while favoring established players. The way forward necessitates a unified worldwide approach—one that guarantees fair remuneration, acknowledges complex authorship, and encourages innovation without compromising creator rights. Although laws like the Music Modernization Act and the Protect Working Musicians Act are encouraging first steps, more thorough and coordinated approaches are needed to safeguard both human and hybrid forms.

*******


* Student, The National University of Advanced Legal Studies, India. Email id: sjayavardhini2167@nuals.ac.in.

[1] The Future of Music Intellectual Property: Trends and Insights - IP Law Mastery

[2] Copyright Laws and Regulations India 2025 | LexOrbis

[3] TRIPS Agreement

[4] WIPO Copyright Treaty (WCT) (1996) :

[5] Musical_Work.pdf

[6] Copyright Laws and Regulations India 2025 | LexOrbis

[7] Section 22 in The Copyright Act, 1957

[8] Copyright and Artificial Intelligence, Part 2 Copyrightability Report

[9] Recent Developments in AI, Art & Copyright: Copyright Office Report & New Registrations - Center for Art Law

[10] Burrow-Giles Lithographic Co. v Sarony, 111 U.S. 53 (1884), 58.

[11] Case Review: Thaler v. Perlmutter (2023) - Center for Art Law

[12] AI created a song mimicking the work of Drake and The Weeknd. What does that mean for copyright law? - Harvard Law School | Harvard Law School

[13] No Human Authorship, No Copyright | TALG®

[14] Case Studies: Legal Battles Over AI-Generated Music

[15] [2003] EWHC 1315 (Ch) (Eng.).

[16] 1978 4 SCC 118

[17] 1996 SCC ONLINE KER 63

[18] digitaldollar_execsummary.pdf

[19] The Impact of Streaming on Content Licensing: Challenges and Opportunities - Vitrina

[20] Streaming, Music, and Rights Management • Rightsline

[21] AI Copyright Law 2025: What Creators Must Know – Jack Righteous

[22] Randy Travis releases first song post-stroke with help of AI - Los Angeles Times

[23] AI Music Outputs: Challenges to the Copyright Legal Framework

[24] Microsoft Word - 13. V.K. Ahuja.docx

[25] Music AI Copyright Law Global Update | Studio Legal

[26] Music AI Copyright Law Global Update | Studio Legal

[27] Copyright Office

[28] Al-Generated Music and Copyright- India - S.S. Rana & Co.

[29] New bill aiming to give artists "at least" 1 cent per stream put before US Congress - News - Mixmag

[30] untitled

[31] Congresswoman Ross Introduces Legislation to Support Independent Musicians and Ensure Fair Negotiations with Streaming Platforms and AI Developers | Press Releases | Representative Deborah Ross

[32] Understanding the Music Modernisation Act: copyright reform in the digital age - iPleaders