Introduction
Artificial Intelligence has made something extraordinary and unsettling. It can now clone a human voice so perfectly that even trained listeners struggle to tell the difference. Two moments in recent years made this challenge visible to ordinary audiences. The first was global: the viral AI-generated song Heart on My Sleeve, which sounded exactly like a Drake and The Weeknd collaboration, even though neither artist had recorded a single note. The second was closer home: an AI-generated version of the Bollywood song Saiyaara, recreated in the unmistakable voice of Kishore Kumar, which left Indian listeners torn between nostalgia and discomfort.
Both incidents capture the same core problem: when AI can duplicate a singer’s voice, it interferes with identity, legacy and livelihood. But while many countries are still debating how to respond, India’s legal framework, through constitutional rights, copyright principles, personality rights, digital regulations, and proactive courts, already offers quite meaningful protection. Understanding how this works requires a closer look at Indian copyright law and how courts are interpreting it in the age of AI.
Background: Copyright, Performers’ Rights and Digital Protection in India
At first glance, copyright law may seem limited in dealing with AI-generated voice clones. A voice itself is not copyrighted. However, the Copyright Act, 1957 (the Act), contains several provisions that work together to protect the creative expression behind that voice.
Section 14 of the Act sets out the exclusive rights of authors and owners over copyrighted works. Although it primarily focuses on compositions and sound recordings, it creates the legal environment in which a performer’s identity gains recognition and commercial value.
More directly, sections 38 and 38A grant performers’ rights. A singer’s performance, i.e., their vocal expression, tone, delivery and on-stage or studio presence, belongs to them and no one may record, reproduce or communicate such a performance without permission. Courts have interpreted these rights broadly, especially as new technologies emerge. But when an AI system generates a new song and is fed a cloned voice that was engineered to match the artist’s distinctive vocal texture, we may dangerously approach the line where “new” performances are created without consent. That reading is supported by section 2(q), which treats “any acoustic presentation of a singer” as something falling within the meaning of a “performance.” So if an A.I. model comes up with a new recording in some fashion, but its purpose is to sound like the original vocal persona, that would also be protected under these sections. Beyond copyright, India also has a surplus in its digital laws. The IT Act differentiates between false and injurious content, and deals with these variations. Section 66E deals with the violation of privacy, while section 67 is related to harmful digital content, and section 69A empowers the government to block any information that could mislead, be prone to harm or threaten the privacy and dignity of any person. The onus is once more on social media platforms and intermediaries to act swiftly against distorted and manipulated content, including deepfakes, under the IT Rules, 2021. Platforms are legally responsible for exercising due diligence, removing illegal content at the earliest and acting on user complaints not later than a fixed time-frame. Those that don’t run a risk of losing their safe-harbour protections. Realising the looming threat of synthetic media, the Ministry of Electronics and Information Technology (MeitY) had also issued various advisories against the misuse of deepfake which includes direction to watermark or label the AI-generated content, setting up mechanisms for removal and correction, etc. The Government of India has even set up a special committee to establish rules for AI-generated content, deepfake prevention and digital impersonation by 2025. This is a steered process towards putting in place a future-proof regulatory framework.
Main Discussion: How Indian Courts are Shaping Voice Protection
The law provides a solid base, but the true change has been driven by court decisions that explicitly deal with the question of identity in the age of AI. Character Rights in India Over the last few years, Indian courts have been steadily extending personality rights to adapt them to the demands of digital technologies and AI. What was originally a protection for names, images, and likenesses has now matured into the concept of recognising ‘voice’ as a separate and legally protectable aspect of identity. A chronology of recent decisions shows a clear and deliberate trajectory.
The turning point came in 2023, when the Delhi High Court decided Anil Kapoor v. Simply Life India. Faced with the unauthorised use of Kapoor’s name, likeness, catchphrases, and voice in merchandise and AI-generated content, the Court acknowledged the growing capacity of technology to distort an individual’s public persona. Crucially, it regarded those elements also as forms of human dignity and livelihood under Article 21 of the Constitution. By issuing a wide-ranging injunction, the Court indicated that Indian jurisprudence sees identity not as splintered elements to be picked from here and there but as an integrated whole that deserves constitutional protection.
This line of reasoning was crystallised in 2024 by the Bombay High Court in Arijit Singh v. Codible Ventures LLP. The defendant ran an AI-tool which enabled users to convert their own singing voice into a version of Singh’s AI-created Voice. And although the company claimed that PROPHECY copied no single song, this formalistic defence was discredited by the Court. It concluded that Singh’s voice possessed a deep reservoir of artistic and commercial worth, and there was a genuine risk of consumer confusion regarding AI-generated impersonation. The opinion recognised that voice as such, stripped of any association with a particular song, is at the heart both of artistic identity and economic opportunity.
The jurisprudence was expounded upon in 2025 when the Delhi High Court addressed AI-generated content (which polluted cyberspace and constituted infringement) using Nagarjuna Akkineni’s voice without permission in Nagarjuna Akkineni v. Various Online Platforms. The Court confirmed that a celebrity’s vocal identity is a valuable asset for commerce and reputation, which deserves imminent protection from digital abuse. Around the same time, in Abhishek Bachchan v. The Bollywood Tee Shop & Ors., 2025, the Court read down unauthorised commercial exploitation of Bachchan’s personality by invoking voice as an integral part of personality rights (along with name and image). The Court regarded the issue as a type of deception, akin to passing off, that would have the public infer there is a connection between the AI-created content and an individual respected by society. It concluded that while technology has since advanced, the law must always safeguard the fundamental constituents of an individual’s personality.
Other than these instances, the Bombay High Court in Asha Bhosle v. Mayk Inc. & Ors., 2025 grappled with voice-based persona rights in particularly clear terms. It also implicitly limited unlicensed AI cloning and commercial exploitation of the singer’s voice, recognising that a person’s voice is inextricably bound up with their identity, reputation and livelihood. Focussing on the unquantifiable damage from voice imitation, the Court enjoined and ordered platforms to delist infringing content, reiterating that in a generative AI world, voice is independently protected under law as an essential characteristic of personality right.
Conclusion: Where India Stands After Kishore Kumar’s “Saiyaara” Moment
The sudden viral popularity of the AI-generated Saiyaara in the voice of Kishore Kumar shows us that we are already on our way into the era of synthetic performances, set registration and more. Impressive, to be sure, for many. But for other people, it seemed to be an invasion of the legacy of one artist whose music defined generations. It provoked ethical and legal questions about consent, cultural memory, and creative authenticity. The good news is that the Indian law was not issued unprepared. The courts have read other voices as an attribute of dignity and identity in privacy-based personality law, and the Copyright Act performer rights protect the economic and expressive worth of the vocal performance. The IT Act, the IT Rules and MeitY’s deepfake guidelines impose even more explicit responsibility on platforms and law enforcement authorities to act with alacrity against impersonation. This latter trend has been even more strongly emulated at the policy level. In December 2025, the Department for Promotion of Industry and Internal Trade (DPIIT) circulated a working paper on Generative AI and Copyright, recognising that artificial intelligence is a game-changer for creative industries and sending out a warning against unfettered use of any work capable of being produced. The paper marks a shift towards structured licensing, accountability and equitable compensation architectures vis-à-vis commercial exploitation, having regard to the manner in which regulation should strike a balance between fostering creativity and protection of human intellect. The big lesson is that as AI gets better at creating performances of practically human quality, the law must protect not just works, but creators. A singer’s voice is more than a soundwave; it carries emotion, memory, skill and identity. Whether it belongs to legends like Kishore Kumar or contemporary stars like Arijit Singh, it deserves protection against unauthorised replication.
In the years ahead, India will likely refine these protections further. But even today, the country offers a strong, flexible and human-centred legal framework, one that tells AI creators and platforms the same simple message: you may innovate, but you cannot impersonate.
References
- Harvard Law School, AI Created a Song Mimicking the Work of Drake and The Weeknd: What Does That Mean for Copyright Law?
https://hls.harvard.edu/today/ai-created-a-song-mimicking-the-work-of-drake-and-the-weeknd-what-does-that-mean-for-copyright-law/ - NDTV, “Saiyaara” Song Recreated in Kishore Kumar’s Voice Using AI, Internet in Awe
https://www.ndtv.com/offbeat/saiyaara-song-recreated-in-kishore-kumars-voice-using-ai-internet-in-awe-too-good-8963459 - The Hindu, Decoding Personality Rights in the Age of AI
https://www.thehindu.com/opinion/op-ed/decoding-personality-rights-in-the-age-of-ai/article70322498.ece - Copyright Act, 1957, s. 14
- Copyright Act, 1957, ss. 38 & 38A
- Copyright Act, 1957, s. 2(q)
- Information Technology Act, 2000, s. 66E
- Information Technology Act, 2000, s. 67
- Information Technology Act, 2000, s. 69A (Blocking public access to information)
- Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021, as amended
- Ministry of Electronics and Information Technology (MeitY), Advisory on Deepfakes and Misinformation (2025)
https://www.meity.gov.in/static/uploads/2025/10/8e40cdd134cd92dd783a37556428c370.pdf - Anil Kapoor v. Simply Life India & Ors., Delhi High Court (2023)
- Constitution of India, art. 21
- Arijit Singh v. Codible Ventures LLP, Bombay High Court (2024)
- Nagarjuna Akkineni v. Various Online Platforms, Delhi High Court (2025)
- Abhishek Bachchan v. The Bollywood Tee Shop & Ors., Delhi High Court (2025)
- Asha Bhosle v. Mayk Inc. & Ors. Interim Application (L) No. 30382 of 2025 in Commercial IP Suit (L) No. 30262 of 2025 (Bombay High Court)
- SSRana & Co., Personality Rights and the Bachchan Cases: A New Chapter in Indian Jurisprudence
https://ssrana.in/articles/personality-rights-and-the-bachchan-cases-a-new-chapter-in-indian-jurisprudence/ - Department for Promotion of Industry and Internal Trade (DPIIT), Working Paper on Generative Artificial Intelligence and Copyright (2025)
https://www.dpiit.gov.in/static/uploads/2025/12/ff266bbeed10c48e3479c941484f3525.pdf
