AI, Deepfakes and Personality Rights in India: Legal Framework and Key Judicial Trends 

Posted On - 5 December, 2025 • By - Aurelia Menezes

With key insights from the recent Bombay High Court order in Asha Bhosle v. Mayk Inc. & Ors. 

Introduction

Rapid advances in artificial intelligence particularly deepfake technology and voice-cloning have challenged traditional doctrines governing privacy, personality rights, and intermediary liability. Indian courts are increasingly being called upon to determine whether a celebrity’s voice, likeness, or other attributes can be commercially exploited without consent. 

The Bombay High Court’s ad-interim order in Asha Bhosle v. Mayk Inc. & Ors. adds an important layer to this evolving jurisprudence. The ruling, while specific to the facts of the case, reflects broader legal questions surrounding AI-generated impersonation, platform liability, and cross-border enforcement. This article examines the broader framework and uses the Bhosle case as a key illustration, alongside other relevant precedents. 

Personality Rights in India: Scope and Nature

India does not have a stand-alone “Right of Publicity Statute,” but courts have consistently held that a celebrity’s commercially valuable persona including name, likeness, voice, signature traits, and distinctive attributes are all protectable under: 

  • Right to Privacy (Article 21) 
  • Tort of Passing Off 
  • Unfair Trade Practices under Consumer Protection law 
  • Intellectual Property principles, particularly where goodwill and misrepresentation coexist 

Key jurisprudence: 

  • R. Rajagopal v. State of Tamil Nadu (1994) – foundational privacy principles. 
  • ICC Development International v. ARVEE Enterprises (Delhi HC, 2003) – publicity rights arise from an individual’s persona and cannot be commercially exploited without consent. 
  • Titan Industries v. M/s Ramkumar Jewellers (Delhi HC, 2012) – unauthorized use of Amitabh Bachchan and Jaya Bachchan’s images held to be passing off; publicity rights recognized as commercial assets. 
  • Aishwarya Rai Bachchan v. Aishwaryaworld.com (Delhi HC, 2009) – held that a celebrity’s name itself carries goodwill. 
  • DM Entertainment v. Baby Gift House (Delhi HC, 2010) – unauthorised sale of “Daler Mehndi dolls” constituted infringement of personality rights. 

These cases collectively establish that a celebrity’s persona is a proprietary right, enforceable against unauthorized commercial use. 

AI tools can replicate not only likeness but also voiceprints, arguably the most distinctive component of a singer’s identity. Voice-based misappropriation raises: 

  • Privacy violations 
  • Economic loss due to misappropriation of goodwill 
  • Consumer deception 
  • Reputational injury, especially when deepfakes depict activities never undertaken 

Thus, courts must determine whether voice, an intangible biomarker, can be exclusively owned for commercial purposes. The Bombay High Court affirmed this principle in Arijit Singh v. Codible Ventures LLP (2024), restraining AI-generated mimicry of Singh’s voice and data scraping of his videos. 

Intermediary Liability under the IT Act and E-commerce Rules

Intermediaries generally enjoy safe harbour protection under: 

  • Section 79, IT Act 
  • Intermediary Guidelines (2021) 
  • Consumer Protection (E-commerce) Rules, 2020 

However, this protection is contingent on: 

  • absence of knowledge, 
  • prompt takedown upon notice, 
  • non-participation or control in the infringing activity. 

Judicial trend (e.g., MySpace Inc. v. Super Cassettes, Delhi HC 2017) indicates that once notified, intermediaries have an affirmative duty to act diligently. 

Case Illustration: Asha Bhosle v. Mayk Inc. & Ors. (Bombay HC, 2025)

1. Facts & Allegations 

The plaintiff alleged: 

  • AI platform Mayk Inc. enabled training and generation of songs in her voice. 
  • YouTube deepfake videos depicted her in fabricated scenarios. 
  • Several sellers sold “Asha Bhosle merchandise” without permission. 
  • Large intermediaries (Amazon, Flipkart, Google) hosted or monetised infringing content. 

2. Issues 

The dispute raised broader legal questions: 

  • Does Indian law protect voice as a personality attribute? 
  • Can courts restrain offshore AI platforms accessible in India? 
  • What duties do e-commerce and content-sharing intermediaries owe? 
  • Can artistic expression or fair use justify AI-based impersonation? 

3. Court’s Findings and Orders 

The Bombay High Court held that: 

1. Voice is an integral part of personality rights: Relying on precedents like Aishwarya Rai Bachchan, the Court held that misappropriation exists even without misrepresentation; mere unauthorized exploitation is actionable. 

2. AI voice-cloning constitutes prima facie infringement: Converting the plaintiff’s original recordings into AI-generated outputs that imitate her is a “brazen appropriation of the plaintiff’s labour and reputation.” The Court likened voice cloning to creating new copyrighted works falsely attributed to the celebrity. 

3. Intermediary safe harbour is not absolute: Once platforms were notified, they were required to: 

  • remove infringing listings, 
  • disclose subscriber information, 
  • prevent re-uploading, 
  • delist unauthorised merchandise. 

4. Jurisdiction extends to foreign platforms: Because harm occurred in India and services targeted Indian users, offshore presence was held irrelevant at the interim stage. 

5. Broad ad-interim injunction: The Court restrained use of name, image, likeness, voice, and any deceptively similar indicia. AI voice tools were ordered to be suspended temporarily. 

1. Expansion of Personality Rights into the AI Era: This ruling reinforces judicial recognition that: 

  • personality rights are not restricted to physical likeness; 
  • voice is equally proprietary and protectable; 
  • AI systems cannot rely on claims of innovation to bypass consent requirements. 

The Court also aligned with global trends (e.g., U.S. cases involving Tom Waits, Bette Midler), where unauthorised voice imitation was held actionable. 

2. Deepfakes as a Legal Category: The order implicitly classifies deepfakes as misleading representations, possibly defamatory, and violative of privacy and publicity rights. This opens space for future legislation regulating synthetic media. 

3. Intermediary Obligations After Notice: Courts are adopting a stricter view: safe harbour is not a “blank cheque”. Platforms that profit from infringing content or ignore notice can face liability, as seen in earlier Christian Louboutin SAS v. Nakul Bajaj (Delhi HC, 2018), Kent RO Systems v. Amit Kotak (Delhi HC, 2017). This decision strengthens that trajectory. 

4. Cross-Border Enforcement and Long-Arm Jurisdiction: Courts are increasingly willing to restrain foreign entities, compel disclosure, and require takedown of India-targeted content. This harmonises with earlier directions in Swami Ramdev v. Facebook/Google/Twitter (Delhi HC, 2019). 

Conclusion

The Asha Bhosle v. Mayk Inc. order marks a significant evolution in India’s approach to AI-generated impersonation and synthetic media. It confirms that: 

  • voice, like name and image, is a protectable personality right; 
  • AI-driven misappropriation will be treated as passing off and privacy infringement; 
  • intermediaries share proactive responsibility once notified; 
  • foreign platforms are accountable where harm is suffered domestically. 

While the final ruling is awaited, this decision sets the groundwork for future disputes on voice cloning, deepfakes, and digital identity, and underscores the need for a dedicated statutory framework to regulate AI-generated content.