AI-generated replicas of deceased persons: Examining Posthumous Personality Rights in India

Abstract

As artificial intelligence technologies advance rapidly, synthetic media capabilities such as voice cloning and deepfakes enable the creation of realistic digital replicas of deceased individuals. Termed “digital resurrection,” this process facilitates commercial, recreational, and personal applications, allowing the preservation of memories and cultural heritage. However, it raises significant legal and ethical questions regarding personal autonomy, human dignity, privacy, and unauthorized exploitation. While Indian courts have increasingly recognized personality rights for living individuals, the jurisprudential status of posthumous personality rights remains unsettled. India currently lacks a dedicated statutory framework to regulate AI-generated replicas of deceased persons, leaving personas vulnerable to misappropriation and reputational harm. This article analyzes the technical and legal dimensions of digital resurrection, evaluates existing Indian personality rights jurisprudence, identifies statutory deficiencies, examines comparative legal frameworks from the United States, the European Union, and South Korea, and proposes a comprehensive legislative framework to protect post-mortem personality rights in the digital age.

Keywords: Artificial Intelligence, Cloning, Deepfake, Digital Resurrection, Privacy, Personality Rights, Posthumous Rights, Reputation, Statutory Reform, Digital Governance.

I. Introduction

The year 2023 witnessed developments that would have been considered science fiction a decade earlier: a major Hollywood studio utilized artificial intelligence to reconstruct the voice of a deceased actor for a posthumous film credit, and a global streaming platform released an AI-generated music track featuring the synthetic voice of a recently deceased pop star. Neither entity had obtained explicit consent from the deceased individuals’ estates prior to production. Both incidents ignited debate regarding the boundary between technological tribute and commercial exploitation, a conflict that Indian law is presently ill-equipped to resolve.

The tools enabling these controversies are no longer exclusive to well-capitalized media conglomerates. Voice-cloning software capable of producing a near-perfect sonic replica of an individual requires only a few minutes of audio input and is widely available as a commercial subscription product. Diffusion-based image models can synthesize photorealistic likenesses from a limited set of photographs. Retrieval-augmented generative systems can be fine-tuned on the written output of a deceased person, including letters, interviews, and public speeches, to produce conversational responses that convincingly simulate that individual’s style of thought and expression. Taken together, these technologies enable third parties to resurrect a deceased person in digital form without the knowledge or approval of surviving family members.

India is directly affected by this technological paradigm. The nation’s entertainment industry is among the most prolific globally, and substantial commercial value attaches to the likenesses of deceased film stars, musicians, and public figures. Furthermore, India hosts a rapidly growing technology sector developing generative AI tools, alongside culturally significant communities for whom the digital image of a deceased religious or political figure carries profound emotional resonance. The potential to exploit these digital assets for commercial gain or ideological persuasion is evident, whereas the legal mechanisms to regulate such deployment remain underdeveloped.

This article proceeds in six parts. Part II examines the technological architecture, applications, and legal stakes of digital resurrection. Part III analyzes the doctrinal foundations and judicial development of personality rights under Indian law. Part IV identifies current statutory gaps regarding posthumous protection. Part V examines comparative legal approaches in the United States, the European Union, and South Korea. Finally, Part VI outlines proposed principles for an Indian legislative framework governing post-mortem personality rights in the age of artificial intelligence.

II. Digital Resurrection: Technology, Applications, and Stakes

A. The Technical Architecture of Digital Resurrection
Digital resurrection serves as a broad term for a cluster of AI-driven processes enabling the synthetic recreation of an individual’s physical appearance, voice, movement, and communicative style. Deep neural network-based voice cloning systems identify the spectral and prosodic features characterizing an individual speaker’s voice. Trained on audio corpora as brief as three to five minutes, these models synthesize new speech from arbitrary text inputs. Contemporary commercial platforms such as ElevenLabs, Resemble AI, and Descript Overdub have made these tools accessible to non-specialist users.

Deepfake video generation, transitioning from generative adversarial networks toward diffusion-based architectures, produces temporally coherent, high-resolution video of a target individual’s face and body in arbitrary poses and expressions. Current academic and commercial systems can generate convincing synthetic video from a single still photograph. Combined with voice cloning, the resulting audio-visual artifact can be difficult for lay viewers to distinguish from authentic footage without technical forensic analysis.

Large language model fine-tuning enables textual replication. Pre-trained models fine-tuned on an individual’s documented output, such as published works, personal correspondence, interview transcripts, and social media archives, can generate text matching that person’s register, vocabulary, and style. Unlike visual deepfakes, textual replicas operate in a medium carrying an inherent presumption of authenticity when attributed to a named individual.

B. Domains of Application
The applications of digital resurrection span diverse social contexts. In personal settings, memorialization platforms offer interactive digital avatars of deceased relatives to preserve family legacies. Academic institutions have similarly experimented with AI-assisted oral history archives, enabling interactive queries in a subject’s synthetic voice.

In the entertainment industry, digital resurrection is used to complete media productions following an actor’s death, expanding on earlier computer-generated imagery techniques. However, unauthorized commercial deployments of deceased recording artists’ synthetic voices and likenesses have created intellectual property and estate disputes.

In political and religious spheres, generating speeches attributed to deceased leaders can serve educational purposes or be deployed to fabricate endorsements, revise historical records, or influence public sentiment. In India, given the public reverence for historic figures and iconic cultural personalities, unauthorized AI replications carry implications for social stability and public discourse.

C. Legal Stakes in the Indian Context
Several factors heighten the necessity of legal clarity in India. First, the commercial value associated with iconic performers in major film industries creates economic incentives for unauthorized posthumous exploitation. Second, rapid digital expansion across an internet user base exceeding 800 million accelerates the dissemination of viral synthetic content, making post-hoc legal remedies difficult to enforce effectively. Third, existing statutory frameworks, designed primarily for analogue contexts, have not been configured to address the specific challenges presented by posthumous synthetic replication.

III. Personality Rights Under Indian Law: Doctrinal Foundations and Judicial Development

A. Constitutional Provisions
The Constitution of India does not explicitly articulate a standalone right of personality. However, fundamental rights under Article 21 guarantee the protection of life and personal liberty.[1] In K.S. Puttaswamy v. Union of India (2017), the Supreme Court affirmed that the right to privacy is an intrinsic part of life and personal liberty under Article 21.[2]

The Court held that personal autonomy, dignity, and control over identity are essential components of privacy. Furthermore, courts have read personality rights in conjunction with freedom of speech and expression under Article 19(1)(a) to protect reputation and personal identity.[3] These constitutional interpretations provide the foundation for judicial recognition of personality rights in the absence of an explicit statute.

B. Judicial Development of Personality Rights
1. The Right of Publicity: Indian jurisprudence on personality rights has largely developed through common law rulings addressing the right of publicity, which grants individuals, particularly public figures, control over the commercial exploitation of their name, image, and likeness. In Titan Industries Ltd. v. Ramkumar Jewellers (2012), the Delhi High Court recognized that unauthorized commercial use of a celebrity’s identity constitutes an actionable infringement of personality rights.[4]

Similarly, in Shivaji Rao Gaikwad v. Varsha Productions (2015), the Madras High Court restrained the unauthorized use of actor Rajinikanth’s name, image, and voice, treating personal persona as an entity entitled to legal protection against commercial misappropriation.[5] More recently, in Anil Kapoor v. Simply Life India & Ors. (2023), the Delhi High Court issued an injunction protecting the actor’s name, voice, image, and AI-generated avatars from unauthorized commercial exploitation, explicitly addressing the risks posed by generative AI tools.[6] While these rulings established protections for living individuals, their application to deceased personas remains textually unspecified under current law.

2. Privacy and the Right to Be Forgotten: Following the Puttaswamy decision, judicial discussions have extended to dignity-based interests, including the right to be forgotten. While not codified as an absolute right, several High Courts have issued orders directing the removal of personal data from digital platforms. In the posthumous context, this raises questions regarding control over a deceased individual’s digital trail and data archives utilized to train AI models.

3. Moral Rights Under the Copyright Act: Section 57 of the Copyright Act, 1957 grants authors moral rights, including the right to claim authorship and object to distortion or modification prejudicial to their honor or reputation.[7] Section 57(2) provides that these moral rights may be exercised by the author’s legal representatives post-mortem for the duration of the copyright term.[8] The Delhi High Court affirmed the post-mortem enforceability of moral rights in Amarnath Sehgal v. Union of India (2005).[9] However, Section 57 applies strictly to copyrightable works and their authors, leaving non-author personas, voices, and likenesses outside its statutory scope.

C. Statutory Instruments of Limited Applicability
Other statutory provisions touch upon digital identity indirectly. The Information Technology Act, 2000,[10] alongside the IT (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 (as amended in 2023), obligates intermediaries to address impersonation and synthetic media.[11] These rules require platforms to label synthetic content and remove unauthorized deepfakes upon notification. However, these mechanisms regulate platform compliance rather than establishing substantive posthumous personal rights.

The Digital Personal Data Protection Act, 2023 (DPDPA) establishes a framework for processing personal data.[12] However, its protections apply to living individuals (“data principals”). Upon an individual’s death, statutory data protection rights lapse under the current structure, without clear provisions for estate succession regarding data processed for AI model training.

Additionally, Sections 38 through 38B of the Copyright Act protect performers’ rights for fifty years following a performance.[13] While these rights allow heirs to control recordings of past performances, they do not restrict AI models from generating entirely new synthetic performances that imitate a performer’s style without copying a specific protected recording.

IV. The Legal Vacuum: Why Existing Law Is Inadequate

A. Absence of Statutory Posthumous Personality Rights
The primary deficiency in Indian jurisprudence is the absence of an explicit statute defining personality rights and their posthumous duration. While common law rulings protect living figures, the legal character of publicity rights post-mortem remains unsettled. If viewed as a dignity-based right under Article 21, it encounters the legal principle that personal rights generally terminate at death. If categorized as a property right capable of inheritance, clear statutory authority is required to define its transferability, scope, and term limits.

In the absence of statutory guidance, surviving family members lack explicit mechanisms to restrain technology entities from utilizing a deceased relative’s voice or image as training data, or to claim licensing royalties from commercial synthetic outputs. Establishing claims requires synthesizing causes of action across tort law, equity, and constitutional principles, creating procedural hurdles and cost barriers.

B. Deficiencies in the Consent Framework
Modern data governance relies on valid legal bases, typically requiring explicit consent for processing personal parameters. Under the DPDPA, consent mechanisms apply exclusively to living data principals. Consequently, personal data generated during an individual’s lifetime can be collected from public archives to train generative AI systems without statutory consent requirements from legal heirs.

Furthermore, the enforceability of advance directives executed during an individual’s lifetime regarding the post-mortem use of their data remains uncertain under standard Indian contract law, which generally does not bind an estate post-mortem without specific statutory authorization.

C. Undeveloped Damages Framework
Current tort law remedies focus on quantifiable financial loss or direct personal injury to a living claimant. Quantifying commercial injury to a deceased persona presents evidentiary challenges under traditional tort doctrine. Unlike jurisdictions with established economic valuation models for celebrity personas, Indian civil courts lack standardized valuation methodologies for unauthorized posthumous licensing, limiting the deterrent effect of monetary damages.

D. Limitations of Reactive Intermediary Rules
The safe harbor framework under the Information Technology Act relieves digital platforms of primary liability provided they comply with notice-and-takedown procedures. This reactive structure places the burden on surviving family members to detect unauthorized synthetic content and initiate takedown requests. For non-public figures or families without monitoring resources, reactive regulation offers limited protection against widespread digital dissemination.

V. Comparative Frameworks: Lessons for India

A. The United States
1. State Right of Publicity Statutes: In the United States, post-mortem personality rights are governed primarily at the state level. The California Celebrity Rights Act (1985) explicitly provides that the right to control the commercial use of one’s name, voice, signature, photograph, or likeness survives death for seventy years and is transferable to heirs.[14] Recent amendments explicitly extend protection to digital replicas and synthetic media. In 2024, Tennessee enacted the Ensuring Likeness Voice and Image Security (ELVIS) Act, updating state publicity law to protect an individual’s voice from unauthorized AI generation.[15]

New York revised its Civil Rights Law in 2021 to recognize a statutory post-mortem right of publicity lasting forty years for deceased personalities, incorporating explicit provisions addressing unauthorized digital replicas in commercial media.[16]

2. The Proposed Federal NO FAKES Act: At the federal level, the Nurture Originals, Foster Art, and Keep Entertainment Safe (NO FAKES) Act was introduced to establish a unified federal right protecting individuals against unauthorized AI-generated digital replicas.[17] The proposed legislation establishes a post-mortem right extending between ten and forty years depending on performance status, imposes statutory damages, and establishes liability for producing, hosting, or distributing unauthorized synthetic replicas.

B. The European Union
The European Framework prioritizes fundamental rights and personal dignity over transferable property interests. The General Data Protection Regulation (GDPR) applies exclusively to living natural persons under Recital 27, leaving post-mortem data protection to member state legislation.[18] Member states such as Germany extend dignitary protections post-mortem through general personality rights rooted in constitutional basic rights.

The EU Artificial Intelligence Act (2024) addresses synthetic media by imposing transparency mandates.[19] Providers and deployers of AI systems generating deepfakes or synthetic audio-visual content must disclose that the media has been artificially generated or manipulated, establishing baseline regulatory oversight for AI outputs.

C. South Korea
South Korea regulates synthetic media through data protection laws, industry directives, and unfair competition regulations. In 2023, the Ministry of Culture, Sports and Tourism issued guidelines requiring synthetic media incorporating real individuals’ likenesses or voices to carry clear synthetic labels, classifying unconsented commercial replication as an unfair commercial practice.

D. Analytical Synthesis for the Indian Context
Comparative models provide key structural insights: First, legislative frameworks must reconcile property-based approaches (US model) with dignity-based approaches (EU model). Given that Indian constitutional jurisprudence grounds privacy in dignity while recognizing economic interests under Article 300A,[20] a hybrid statutory approach is appropriate. Second, statutory frameworks must establish clear post-mortem term limits to balance estate rights against public domain access and free expression under Article 19(1)(a). Third, proactive transparency mandates, such as mandatory AI labeling, are necessary to complement civil enforcement mechanisms.

VI. Towards an Indian Legislative Framework for Posthumous Personality Rights

A. Principled Foundations
An Indian statutory framework governing post-mortem personality rights rests on three foundational pillars:

1. Dignity Rationale: Article 21 guarantees human dignity as a core constitutional value.[1] Unauthorized synthetic replication that places an individual in defamatory, sexually explicit, or misleading contexts violates the personal dignity maintained during their lifetime. Extending dignitary protections post-mortem aligns with existing statutory principles under Section 57 of the Copyright Act.[7]

2. Property Rationale: The commercial value of a developed persona represents investment, effort, and public recognition. Unsanctioned commercial appropriation of this value by third-party AI developers constitutes an uncompensated taking of economic interests, conflicting with the principles of Article 300A.[20]

3. Relational Rationale: Surviving family members possess legitimate interests in preserving a deceased relative’s memory, protecting dependants’ financial entitlements, and preventing emotional distress caused by unauthorized digital replications.

B. Proposed Legislative Principles
1. Statutory Codification: The Parliament of India should enact dedicated legislation or amend the Copyright Act to explicitly codify personality rights, defining them to include control over the commercial deployment of one’s name, voice, image, likeness, and synthetic persona. The statute should specify that these rights survive death, lasting sixty years post-mortem for commercial personas (matching the standard copyright term under Section 22 of the Copyright Act)[21] and twenty-five years for non-commercial personas.

2. Order of Succession for Enforcement: The statute should establish a priority hierarchy for exercising post-mortem enforcement rights:
(i) Designated representatives named in a valid advance directive;
(ii) Surviving spouse;
(iii) Children (collectively);
(iv) Parents; and
(v) Siblings (collectively).
For public figures of national significance, a designated statutory body should be granted standing to intervene where public interest or historical accuracy is compromised.

3. Mandatory Consent Requirements: Creating, distributing, or commercially exploiting an AI-generated digital replica of a deceased person without written authorization from the statutory successor should be defined as a civil wrong. Valid consent must be preceded by full disclosure regarding technical scope, commercial terms, and intended distribution channels.

4. Statutory Exceptions and Fair Dealing: To preserve freedom of expression under Article 19(1)(a),[3] the framework must incorporate fair dealing exceptions, exempting:
(i) News reporting, commentary, and biographical documentaries;
(ii) Parody, satire, and artistic works with transformative expression;
(iii) Academic, educational, and archival research; and
(iv) Non-personalized computational training of generalized AI models where individual personas are not extracted or commercialized.

5. Judicial Remedies: Statutory remedies should include ex-parte interim injunctions to halt the dissemination of infringing media, compensatory damages based on fair market licensing value, disgorgement of profits derived from unauthorized use, and statutory punitive damages for reckless or intentional violations.

6. Proactive Platform Obligations: AI developers should be required to maintain records of training datasets utilized for persona replication models. Intermediaries hosting user-generated content must implement technical measures, such as automated watermarking detection, metadata verification, and mandatory synthetic media labeling, as prerequisites for safe harbor protections under Section 79 of the IT Act.[22]

7. Amendments to Data Protection Law: The DPDPA should be amended to enable individuals to designate a “data nominee” authorized to exercise data rights post-mortem, including requesting data erasure and restricting data processing for AI model training for up to twenty-five years following the data principal’s death.

VII. Conclusion

The advent of digital resurrection technology presents novel legal challenges at the intersection of artificial intelligence, intellectual property, and constitutional rights. While Indian jurisprudence has recognized personality rights for living individuals through common law developments, the lack of explicit statutory provisions governing post-mortem digital replicas leaves significant regulatory gaps. Unsanctioned AI-generated cloning risks compromising personal dignity, commercial value, and familial interests.

Addressing these challenges requires structured legislative action. By combining property and dignity protections, establishing clear consent mechanisms, defining term limits, providing fair dealing exceptions under Article 19(1)(a), and mandating proactive platform compliance, India can establish a balanced regulatory framework. Such legislative reform is necessary to ensure that technological progress respects individual identity, human dignity, and the integrity of the digital afterlife.

References

[1] INDIA CONST. art. 21.
[2] K.S. Puttaswamy v. Union of India, (2017) 10 SCC 1 (India).
[3] INDIA CONST. art. 19, cl. 1(a).
[4] Titan Industries Ltd. v. Ramkumar Jewellers, 2012 SCC OnLine Del 2382 (India).
[5] Shivaji Rao Gaikwad v. Varsha Productions, 2015 SCC OnLine Mad 158 (India).
[6] Anil Kapoor v. Simply Life India & Ors., CS(COMM) 652/2023 (Del. HC Sept. 20, 2023) (India).
[7] Copyright Act, No. 14 of 1957, § 57, INDIA CODE (1957).
[8] Copyright Act, No. 14 of 1957, § 57(2), INDIA CODE (1957).
[9] Amarnath Sehgal v. Union of India, (2005) 30 PTC 253 (Del. HC) (India).
[10] Information Technology Act, No. 21 of 2000, INDIA CODE (2000).
[11] Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 (India).
[12] Digital Personal Data Protection Act, No. 22 of 2023, INDIA CODE (2023).
[13] Copyright Act, No. 14 of 1957, §§ 38–38B, INDIA CODE (1957).
[14] CAL. CIV. CODE § 3344.1.
[15] Ensuring Likeness Voice and Image Security (ELVIS) Act, TENN. CODE ANN. § 47-25-1101 (2024).
[16] N.Y. CIV. RIGHTS LAW § 50-f.
[17] Nurture Originals, Foster Art, and Keep Entertainment Safe (NO FAKES) Act, S. 2281, 118th Cong. (2023–2024).
[18] Regulation (EU) 2016/679 (General Data Protection Regulation), rec. 27, 2016 O.J. (L 119) 1.
[19] Regulation (EU) 2024/1689 (Artificial Intelligence Act), 2024 O.J. (L 1689) 1.
[20] INDIA CONST. art. 300A.
[21] Copyright Act, No. 14 of 1957, § 22, INDIA CODE (1957).
[22] Information Technology Act, No. 21 of 2000, § 79, INDIA CODE (2000).