top of page
Vrinda Sehgal & Divanshi Gupta

It’s Time to Voice Out: AI Voice Cloning and Misuse

Introduction


In the dynamic confluence of art and technology, Artificial Intelligence (AI) stands as a transformative force. As it evolves, AI not only redefines our visual understanding by recreating portraits and historical events but also extends its influence into the auditory realm, reshaping the way we perceive and engage with sound. The advent of voice cloning technology has added a new dimension, causing a whirlwind in the digital world and emphasizing the far-reaching impact of AI across diverse realms. Many celebrities, including actor Emma Watson, Prime Minister Narendra Modi, and celebrity chef Anthony Bourdain, have fallen victim to voice cloning and its use in "deepfake" videos. More recently, the voice of American rapper Jay-Z was cloned and used to create a YouTube video without his consent. Roc Nation, Jay-Z’s agency, sought a takedown order against YouTube on grounds of copyright infringement of this "deepfake" video. This situation raises legal and ethical concerns surrounding the misuse of voices for deceptive purposes and has sparked debate worldwide over the limits of artificial intelligence, especially in terms of voice cloning.


The use of artificial intelligence for voice cloning, also known as voice synthesis, involves creating a replication of a person’s voice through a completely computer-generated system. This technology has significantly benefited various industries, including audio books, voice assistants, and voiceovers. For instance, in the creation of the Obi-Wan Kenobi TV series, James Earl Jones signed off on allowing Disney to replicate his vocal performance as Darth Vader in future projects using an AI voice-modeling tool called Respeecher.


However, the legal implications surrounding AI-generated voices have come under scrutiny. Scammers or cybercriminals use AI-generated voices to mimic individuals for malicious purposes and to exploit victims by demanding funds or engaging in other fraudulent transactions. Therefore, even staunch proponents of liberal ideologies find such applications of artificial intelligence concerning, not only in terms of invasion of privacy but also as a violation of intellectual property rights. 


Copyright Over a Person’s Voice


In order to be copyrightable, a person’s voice must be fixed in a tangible medium of expression. This implies that the person’s voice must be recorded or captured in some form. In the case of Midler v. Ford Motor Co., the United States Court of Appeal held that there can be no copyright over a person’s voice.  The Court of Appeal held that a voice is not copyrightable per se in the Copyright Act, but common law rights could be enforced since a voice is as distinctive as one’s face. Further, even in the case of Butler v. Target Corp., the United States District Court found that while lyrics to a song are copyrightable, the underlying voice is not.  In both cases, it was reiterated that public figures have the right to protect their voice against misuse for commercial gain under their publicity rights and the right to privacy. 


Recently, even in the case of Amitabh Bachchan v. Rajat Nagi and Ors., the Delhi High Court passed an interim order to prevent the unlawful use of Bollywood megastar Amitabh Bachchan's name, image and voice. 


As per the doctrine of fair use, a person’s voice may be used for commentary or criticism or education even without violating any copyright protection. In addition, a person may also waive copyright protection under a license or assignment agreement. Therefore, copyright protection for a person’s voice under existing copyright laws may be difficult to claim and perhaps some reform with expansion in the application of copyright protection may be a suggested way forward.


False Representation


Unauthorized emulation of voices, whether an artist's or others, using AI technology raises significant concerns related to the tort of "passing off." In the case of Waits v. Frito-Lay, the Court affirmed a celebrity's right to protect a distinctive voice from commercial misappropriation and upheld the viability of voice misappropriation as a tort. The Court further clarified that for a voice to be misappropriated, it must be distinctive, widely known and deliberately imitated for commercial use.  In addition to "passing off," victims of voice cloning can potentially initiate defamation claims against the entity responsible for generating the clone, further emphasizing the multifaceted legal challenges associated with the misuse of voice cloning technology.


Privacy Concerns


The advent of voice cloning technology has thrust privacy concerns into the limelight, offering a double-edged sword of innovation and vulnerability. This cutting-edge capability enables the uncanny replication of an individual's voice, laying the groundwork for a host of malicious activities such as impersonation and unauthorized access. The intricate mechanics of AI voice cloning involve the meticulous collection and preprocessing of data, a process that, when exploited by malicious actors, can lead to legal consequences. Voice recordings initially gathered for legitimate purposes may inadvertently become instruments for illicit activities, ranging from fraud to potential blackmail. Negotiating this legal landscape calls for stringent regulations to govern the ethical use of voice cloning technology, emphasizing the importance of legal frameworks to safeguard individuals from the potential misappropriation of their vocal identities.


Conclusion


In India and globally, the legal framework for artificial intelligence, particularly concerning voice cloning, is unclear. Existing laws lack precision on intellectual property rights for voices, leaving room for misuse like "deepfakes." Privacy torts inadequately cover AI aspects, necessitating a comprehensive legal framework. Addressing ethical and legal concerns requires new regulations harmonizing with existing laws. Given the borderless nature of the digital world, international cooperation is crucial to effectively safeguard against potential abuses. A unified global effort is essential for ensuring responsible and ethical use of voice cloning technology. 



References:

 Midler v. Ford Motor Co., 849 F.2d 460 (9th Cir. 1988)

 CS (COMM) 819/2022




Comments


Featured Posts
Recent Posts
Archive
Search By Tags
Follow Us
  • Facebook Basic Square
  • Twitter Basic Square
  • Google+ Basic Square
bottom of page