Deepfakes, Voice Cloning, and the Right of Publicity

An Analysis under Indian Intellectual Property Law

Author: Gopikrishna P R | Institution: ABBS SCHOOL OF LAW, Karnataka State Law University

Abstract

The rapid development of artificial intelligence (AI) has made it possible to create realistic digital representations of individuals through deepfakes and voice-cloning technologies. Although these technologies have legitimate applications in entertainment, education, accessibility, and creative industries, their unauthorised use may result in identity theft, reputational damage, privacy violations, and commercial exploitation. In India, protection of a person’s identity is addressed through personality-rights jurisprudence, constitutional privacy principles, copyright law, and information technology legislation. The Delhi High Court’s decision in Anil Kapoor v. Simply Life India (2023) is particularly significant because it addressed unauthorised use of an actor’s name, image, voice, and other identifiable personality attributes through AI and deepfake technologies. This article examines the Indian legal framework, the relationship between publicity rights and IPR, and the challenges of consent, enforcement, and freedom of expression. It argues that a clearer legal framework is needed while preserving legitimate artistic and technological expression. Keywords: Deepfakes; voice cloning; right of publicity; personality rights; artificial intelligence; copyright; Indian intellectual property law.

1. Introduction

Artificial intelligence has transformed the way images, videos, and audio recordings are created and distributed. Generative AI tools can reproduce a person’s facial appearance, imitate their voice, and generate statements or performances that the person never made. Such content is commonly associated with deepfake technology and AI-based voice cloning. These developments raise legal questions concerning ownership and protection of human identity. If an AI system reproduces an actor’s voice to advertise a product without permission, the act may affect commercial interests, reputation, and control over identity. A fabricated video may also cause reputational or personal harm. Traditional intellectual property laws protect creative expression, inventions, and commercial identifiers. A person’s voice, likeness, mannerisms, and overall persona do not always fit neatly within these categories. The challenge is to determine how existing Indian laws can protect individuals against unauthorised appropriation of identity. Research question: Whether the existing framework of personality rights, copyright, and information technology law in India provides adequate protection against unauthorised AI-generated replicas of an individual’s voice and likeness?

2. Understanding Deepfakes and Voice Cloning

2.1 Deepfakes

Deepfakes are digitally generated or manipulated images, videos, or audio recordings that can make a

lperson appear to say or do something that never occurred. They commonly rely on machine-learning techniques and generative AI models. The technology is not inherently unlawful; it may be used for filmmaking, dubbing, satire, or education. Legal concerns arise when it is used to deceive, impersonate, defame, exploit, or otherwise violate rights.

2.2 Voice cloning

Voice cloning uses AI systems to reproduce distinctive characteristics of a person’s voice, including tone, pitch, pronunciation, and speaking style. Unauthorised cloning may be used to create false endorsements, fabricate statements, facilitate fraud, or reproduce an artist’s voice commercially without permission. A voice may have personal and commercial significance. Depending on the facts, an imitation may raise personality-rights, privacy, copyright, passing-off, or criminal-law issues. Not every imitation automatically infringes all these rights.

3. Meaning and Scope of the Right of Publicity

The right of publicity is the right to control, and in appropriate circumstances commercially exploit, identifiable aspects of one’s personality, such as name, image, likeness, voice, signature, and other distinctive attributes. Personality rights may also encompass dignity, privacy, and protection against misleading or degrading representations; publicity rights focus particularly on commercial appropriation. In India, publicity rights have developed primarily through judicial interpretation rather than through a single comprehensive statute. In Justice K.S. Puttaswamy (Retd.) v. Union of India (2017), the Supreme Court recognised privacy as a fundamental right under Article 21 and other provisions of Part III of the Constitution. The decision did not create a standalone statutory right of publicity, but it provides important constitutional context for disputes involving misuse of identity. Important distinction: Publicity rights are not a blanket ownership right over every reference to a person. Freedom of expression, news reporting, criticism, parody, and other legitimate uses must also be considered.

4. Legal Framework in India

4.1 Constitution of India

Article 21 protects life and personal liberty. The privacy principles recognised in Puttaswamy may be relevant where a deepfake exposes private information, fabricates intimate content, or misrepresents an individual harmfully. Article 19(1)(a), protecting freedom of speech and expression, is also relevant. Article 21 does not automatically establish a claim for every unauthorised imitation of a voice or image.

4.2 Copyright Act, 1957

The Copyright Act protects original works, cinematograph films, sound recordings, and performers’ rights. Section 38A provides specified exclusive rights in relation to performances, including certain rights concerning recording and reproduction. Section 38B protects specified moral rights of performers, including identification and protection against certain prejudicial distortions or modifications. These provisions may be relevant when a deepfake uses or manipulates an actual protected performance. Copyright does not automatically grant ownership of a person’s natural voice. If an AI tool clones a singer’s voice, the analysis may depend on whether protected recordings or performances were used and whether the resulting content falsely suggests endorsement or commerciallyappropriates the singer’s persona.

4.3 Information Technology Act, 2000 

Section 66C: Identity theft involving fraudulent or dishonest use of another person’s electronic signature, password, or other unique identification feature.

 Section 66D: Cheating by personation using a communication device or computer resource.

 Section 66E: Violation of privacy through specified acts involving images of a person’s private areas without consent.

Sections 67 and 67A: Electronic publication or transmission of obscene and sexually explicit material, where statutory requirements are met.

Section 69A: Blocking public access to specified online information in accordance with statutory procedure. These provisions do not criminalise every deepfake merely because it is AI-generated. The relevant statutory elements must be established.

4.4 Intermediary Rules and synthetic content

The Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021, regulate intermediary obligations, including due diligence and handling unlawful content. Amendments concerning synthetically generated information may impose requirements such as labelling, traceability, and platform compliance, subject to the exact operative text. Intermediary regulation and personality rights serve different functions. Labelling or removal obligations do not by themselves settle every question of ownership, consent, damages, or commercial exploitation. For academic work, verify the final Gazette notification, effective date, definitions, exceptions, labelling requirements, and removal deadlines.

4.5 Passing off and civil remedies

Passing off may apply where unauthorised use of a person’s identity misrepresents an endorsement, association, or commercial connection and causes or is likely to cause legally recognised harm. Depending on the facts, a person may seek an injunction or damages through appropriate civil proceedings.

5. Landmark Indian Judicial Decisions

5.1 Anil Kapoor v. Simply Life India (2023)

Delhi High Court, CS(COMM) 652/2023, order dated 20 September 2023. The Court considered unauthorised exploitation of Anil Kapoor’s name, image, likeness, voice, mannerisms, and other attributes of his persona. The interim order restrained specified defendants from using these attributes through AI, machine learning, deepfakes, and related technologies in ways that violated his rights.

Relevance: This decision provides judicial support for protecting identifiable personality attributes, including voice, against unauthorised exploitation. It should not be read as a universal prohibition on all AI imitations or parodies

5.2 Amitabh Bachchan v. Rajat Nagi and Others (2022)

Delhi High Court, CS(COMM) 819/2022. The Court granted interim protection against specified unauthorised commercial uses of Amitabh Bachchan’s identity, including uses associated with websites, merchandise, and services. The case illustrates how courts may apply existing principles to commercial appropriation of identity.

5.3 Jaya Amitabh Bachchan v. Various Defendants (2025)

Delhi High Court, CS(COMM) 1194/2025. Proceedings concerned protection for Jaya Bachchan’s name, image, likeness, signature, voice, and other personality attributes, including alleged unauthorised commercial use and AI-generated material. Check the case file for the precise scope and procedural status of the relevant order.

Case material: Delhi High Court case When citing interim injunctions, distinguish findings at that stage from a final determination of every disputed legal issue.

6. How Deepfakes Challenge Indian IPR Law

Synthetic voice advertisement: An AI-generated voice imitates an actor to promote a product without consent. This may raise publicity-rights, passing-off, and potentially copyright issues. Fabricated public statement: A deepfake makes it appear that a public figure said something they never said. Concerns may include misinformation, reputation, privacy, and applicable IT or criminal laws.

AI-generated entertainment: A creator uses a celebrity-like voice in a fictional parody. This requires a nuanced assessment because personality interests may coexist with freedom of expression and creative freedom.

A framework that prohibits every imitation could restrict legitimate expression, while a framework that fails to address deceptive impersonation could leave individuals without effective protection.

7. Major Legal Challenges

7.1 Absence of a comprehensive publicity-rights statute

Personality-rights protection has developed primarily through judicial decisions and existing laws. The lack of a dedicated comprehensive statute can make the scope of protection less predictable.

7.2 Distinguishing identity from expression

A recognisable voice, image, or mannerism does not make every use unlawful. Courts must consider appropriation, misleading association, infringement of protected material, privacy, and legitimate expression.

7.3 Consent and licensing

Licences may not clearly address AI training, voice synthesis, digital replicas, future uses, or sublicensing. Agreements should specify scope, duration, purpose, permitted commercial uses, and termination conditions.

7.4 Attribution and enforcement

Deepfakes can be uploaded anonymously, distributed across jurisdictions, and reproduced after removal. Identifying the creator, preserving evidence, and obtaining effective remedies may be difficult.

7.5 Freedom of expression

Individuals should be able to seek protection against false endorsements and harmful impersonation. News reporting, parody, criticism, and transformative art should not automatically be unlawful merely because they refer to or imitate a recognisable individual.

8. Suggestions for Legal Reform

Clarify personality-rights protection: Define when unauthorised commercial use of a person’s name, voice, likeness, and distinctive attributes gives rise to a remedy.

Establish consent and licensing standards: Encourage agreements governing digital replicas, voice models, AI training, commercial uses, and termination.

Strengthen remedies: Ensure affected individuals can seek effective injunctions and other appropriate relief without unnecessary delay. Improve intermediary compliance: Implement applicable synthetic-content labelling and traceability requirements while preserving privacy and expression.

Protect legitimate creative expression: Safeguard news, criticism, parody, research,education, accessibility, and other lawful uses. Improve public awareness: Provide accessible complaint mechanisms and guidance on preserving evidence and reporting harmful synthetic media. The objective should not be to prohibit AI technology as a whole. The law should distinguish legitimate creative uses from harmful impersonation and unauthorised exploitation.

9. Conclusion

Deepfakes and voice cloning challenge traditional intellectual property boundaries because they can reproduce a person’s identity without necessarily reproducing a conventional copyrighted work. In India, personality-rights jurisprudence, constitutional privacy principles, copyright law, and information technology legislation provide different forms of protection depending on the circumstances. The Delhi High Court’s decision in Anil Kapoor v. Simply Life India demonstrates the relevance of personality rights to AI-generated impersonation and unauthorised commercial exploitation. However, legal remedies must be applied carefully so that identity protection does not become an unrestricted right to control all references to a person. A more predictable framework should clarify consent, commercial exploitation, remedies, and intermediary obligations while preserving legitimate artistic and expressive uses. The central challenge for Indian IPR law is to ensure that technological innovation does not come at the cost of individual dignity, privacy, and control over identity.