Synthesised voices are no longer a novelty. A few minutes of recorded audio produces an uncannily accurate replica — useful for people who cannot speak, for multilingual dubbing, for accessibility. And useful for fraud.
The question neither societies nor legal systems are prepared for: who owns the voice of the future when AI can reproduce voices from the past?
India is seeing a surge in litigation as public figures turn to the courts to protect their names, faces, voices and catchphrases. These disputes sit at the intersection of technology, intellectual property, privacy, free expression and the economics of fame.
The Bachchan litigation
In 2024–25, Jaya Bachchan and Abhishek Bachchan approached the Delhi High Court over unauthorised use of their names, photographs and AI-generated voice impersonations.
The case drew national attention because:
- AI tools and human impersonators used their voices to endorse products and services;
- deepfake versions of Abhishek Bachchan appeared in fraudulent financial advertisements; and
- Jaya Bachchan’s distinctive voice and expressions were recreated for viral videos without consent.
The court recognised that a person’s voice is not just a sound but an intrinsic part of their identity. The ruling reinforced the growing consensus that voice cloning without consent violates dignity, privacy and economic rights, and gave public figures a firmer foundation for addressing deepfake harms.
What is actually being protected
Three concerns drive these petitions:
- False endorsement — that the public will assume the person supports a product or message when they do not;
- Reputational harm, especially through deepfakes, morphed images or GIFs that are embarrassing or offensive; and
- Privacy and livelihood, where unauthorised use invades privacy or affects commercial earnings.
There is an irony worth acknowledging: many of these arguments come from people whose livelihoods depend on being publicly visible. Scholars describe this as the economics of superstars — those already famous using legal systems to reinforce commercial dominance.
The fraudulent financial advertisement is the strongest case for intervention, and it is worth separating from the rest. A deepfake used to induce people to part with money is fraud, harming the public directly and the person incidentally. That is a different thing from a satirical video, and the remedies should not be identical.
The tension with expression
The conflict between celebrity rights and free expression is not new. Indian courts have repeatedly emphasised that news reporting, criticism, parody, satire, art and political commentary must remain protected.
R. Rajagopal v. State of Tamil Nadu established that personal, private information about an individual cannot be published without consent, unless it forms part of public records.
Recent cases extend that to prevent commercial exploitation of a person’s name, face or voice. But much of a public figure’s persona already exists in the public domain, and using publicly available clips or images in non-obscene, non-defamatory contexts — particularly for satire or commentary — should in principle fall within lawful expression.
That line matters more as orders become broader. A personality rights injunction framed widely enough to cover any use of a person’s likeness reaches commentary and criticism that no one intended to restrict.
The ethical question underneath
If a machine can replicate a human voice perfectly, does that voice belong to the algorithm or to the person whose identity it echoes?
The code may be owned by technology firms. The vocal fingerprint that code gives access to is undeniably human. Consent, transparency and control must prevail — without them, misinformation, fraud and character assassination outweigh the creative benefits of synthetic media.
India has arrived at the point where AI blurs the line between creation and imitation, inspiring and impersonating. These cases point toward legislation that protects individuals from exploitation while preserving expression, creativity and public discourse — and personality rights in India remain judge-made, without a statute defining their scope or limits.
Courts may handle these questions in varied ways, but one thing should be clear: the law is not designed to overprotect fame, but to prevent misuse of fame.
Whether human or synthetic, the voice of the future should rest on consent, dignity and fairness — because although technology lets us reproduce a voice, the identity behind it belongs to lived experience.
The takeaways
- Voice is part of identity — the Delhi High Court has said so directly.
- Fraudulent deepfakes are the strongest case for intervention, and differ from satire.
- Parody, criticism and reporting remain protected — orders should not sweep them in.
- Personality rights are judge-made in India — no statute defines their limits.
Frequently asked questions
Is voice protected as part of personality rights in India? Yes — courts have recognised that a person’s voice is an intrinsic part of their identity, and unauthorised cloning violates dignity, privacy and economic rights.
Can I use a public figure’s image for satire or criticism? In principle yes — Indian courts have consistently protected news reporting, criticism, parody, satire and political commentary.
Is there a statute governing personality rights in India? No — personality rights are judge-made, developed through case law rather than defined by legislation.
What makes AI deepfakes different from ordinary impersonation? Their accuracy and scale, and their use in fraudulent advertising — which harms the public directly, not only the person depicted.