Imagine a future where the voice of a legendary singer, long deceased, is brought back to life to sing new songs. With the advent of AI voice cloning, this future is not just a possibility but a looming reality. This piece explores the implications of AI voice cloning in the music industry, particularly focusing on the legal and ethical challenges it poses.
I recently came across an article stating that Arijit Singh aims to patent his voice. While the voice of Arijit Singh may not be patentable, it is not surprising that artists are looking for ways to protect their attributes in light of the AI wave which has taken the music industry by storm.
The Rise of AI in Music
AI has revolutionized many industries, and music is no exception. The introduction of tools like YouTube’s “Dream Track” highlights the growing influence of AI. This tool allows users to create short tracks using the voices of famous artists, showcasing the power and potential of AI in redefining music production. While YouTube has collaborated with several artists like John Legend, Demi Lovato, T-Pain, and Sia, among others, there are several other AI cloning tools which may not have done such collaborations.
Legal Landscape in India
The legal protection of an artist’s voice is a nuanced area. Generally, an artist’s voice is not protectable under intellectual property laws, except in specific instances. For instance, a unique voice element that has been trademarked as a sound mark, or a recorded voice as a performance, which is subject to copyright protection, can be legally safeguarded. Hence, the voice of an artist primarily falls within the realm of personality and publicity rights. This distinction is crucial in understanding the legal ramifications of AI voice cloning.
In India, the legal framework governing AI and personality rights is still in its infancy. Celebrities like Amitabh Bachchan and Anil Kapoor have begun to navigate these waters, primarily in response to deepfake technologies. The Delhi High Court recently granted an ex-parte, omnibus injunction that effectively restrained 16 entities, and the world at large, from utilizing Mr. Anil Kapoor’s name, likeness, image, and employing technological tools such as AI, face morphing, and GIFs for financial gain or commercial purposes. The Delhi High Court delineated instances like parody and satire where free speech in the context of well-known persons may be protected. However, it held that tarnishing, blackening or jeopardising an individual’s personality or its attributes would be illegal. A copy of the order can be accessed here.
The Question of Deceased Artists’ Voices
It is a settled position of law that there is no posthumous application of personality rights in India. Read our detailed post here on whether celebrity rights are inheritable or not in background of the Sushant Singh Rajput case. The Delhi High Court held that the right to Privacy and Right to Publicity are not heritable and do not survive for espousal.
This got me thinking on what is to stop AI tools from cloning voices of late singers like Lata Mangeshkar or Sidhu Moosewala. In the absence of posthumous personality rights, there are no clear legal barriers to prevent such use. However, this raises several ethical and moral questions.
Key Considerations:
- Data Acquisition for AI Training:
AI requires initial data for voice training. This could lead to potential copyright infringement issues if AI developers use existing recordings without proper licensing from music labels.
Quoting an interesting extract of intellectual property expert Louis Tompros, a lecturer on law at Harvard and partner at WilmerHale as published here:
“..there are two categories of questions — an input question and an output question. On the input side, does the training that is required to create these complex AI models implicate copyright? In other words, if I train an AI by having it listen to a whole bunch of music, have I infringed the copyright of the owners of that music if I did it without their consent? Or is that protected in some way by fair use or otherwise under the copyright statute?
Then there’s the output question, which is, if copyright law gives the owner the exclusive right to create a derivative work based on their own prior work, is creating something using AI based on that other work then a derivative work that only the original copyright owner would have the ability to make? In general, music in the style of someone else is not considered a derivative work for purposes of copyright law and is allowed. But where we have machine learning and AI generating the work, it’s an open question as to whether those outputs themselves are protected. So, both the input and output questions are unresolved and complicated”.
These issues are currently being argued in the case against Microsoft and Open AI in the US. It would also be interesting to see the outcome of this case which would certainly pave way for the issues pertaining to copyright infringement in relation to data utilization for training by AI.
- Voice as Property:
Treating a cloned voice as property opens up intriguing legal debates. Could the voice of a deceased artist be subject to succession laws, akin to tangible or intangible property?
- Global Perspective and Litigations
Globally, AI models are trained on data sourced from the internet, often including copyrighted content. This practice has sparked legal controversies in the West. For instance, litigations surrounding the unauthorized use of AI to replicate the voices of living artists have prompted discussions on the boundaries of intellectual property rights. The case of an AI-generated song using a voice similar to Drake’s, which led to claims of infringement by a record label, is a notable example.
In Midler v. Ford Motor Co., 849 F.2d 460 (9th Cir. 1988), Ford produced a commercial for the Mercury Sable, and they wanted to use the Bette Midler song “Do You Want to Dance.” She did not want to give permission for that, so they hired a Bette Midler impersonator and had her perform the song. The Ninth Circuit held that when you have a distinctive voice of a professional singer that is intentionally imitated to sell something, that’s a violation of the right of publicity under state law.
- Opportunity for Music Labels
An intriguing development in the AI voice cloning arena is the potential for music labels, especially those holding copyrights to songs performed by deceased artists, to utilize these voices in their AI tools. This presents a unique opportunity for music labels to revive and reimagine the works of legendary artists in new productions. This could lead to significant commercial success, tapping into the nostalgia and enduring popularity of these artists.
While this presents a lucrative opportunity, it also raises ethical questions about the preservation of an artist’s legacy. Music labels venturing into this territory must consider how the use of a deceased artist’s voice aligns with the artist’s original intent and public image. Collaborations with artists’ estates could form a respectful approach to posthumous music production.
- The Need for a Specific Legislative Framework
The current legal framework is inadequate to address the complexities introduced by AI in the music industry. There is a pressing need for laws that recognize personality and publicity rights in the context of personal attributes, including voice. Legislation must evolve to address the unique challenges posed by AI, ensuring that the rights of artists, both living and deceased, are respected.
- Responsible Use of AI
Beyond legal considerations, there’s a moral imperative to use AI responsibly. The technology holds immense potential for innovation and creativity in music. However, this should not come at the cost of ethical violations or disrespect to the legacy of artists. The music industry, lawmakers, and AI developers must work together to establish guidelines that encourage the positive use of AI while safeguarding personal and artistic integrity.
Conclusion
The advent of AI voice cloning in the music industry opens up a world of possibilities, from creating new music with the voices of deceased artists to exploring uncharted territories in music production. However, it also brings forth significant legal and ethical challenges that need to be addressed. As AI continues to blend with the realms of creativity and legacy, the music industry, along with lawmakers and AI developers, must tread carefully to ensure that this powerful tool enhances the artistic landscape without compromising the integrity and rights of the artists, both past and present.
