In this guide
1. What personality rights are
2. The legal foundations in India
3. How Indian courts have protected them
4. The DPDP Act layer
5. Generative AI changed the maths
6. A practical protection playbook
7. Frequently asked questions
1. What personality rights are
Personality rights (also called publicity rights or image rights) are a person's right to control commercial use of the attributes that identify them: name, image, likeness, voice, signature, and distinctive traits — a catchphrase, a celebration style, a way of speaking. Two ideas sit inside the term: the right of publicity (only you may profit from your identity) and the right to dignity and privacy (nobody may misrepresent or expose you). A brand putting your face on a hoarding without a licence violates the first; a deepfake putting words in your mouth violates both.
2. The legal foundations in India
India has no single personality-rights statute. Protection is assembled from four sources, and understanding all four matters because enforcement usually invokes several at once:
Privacy as a fundamental right. The Supreme Court's nine-judge Puttaswamy judgment (2017) recognised privacy as a fundamental right under Article 21, and with it the individual's control over their own identity — the constitutional bedrock under every modern personality-rights order.
Passing off. The common-law tort that stops one trader misrepresenting a connection with another. When an ad implies a celebrity endorses a product they've never seen, courts treat the persona like unregistered goodwill being traded on.
Copyright and trademark law. Performers' rights under the Copyright Act protect recorded performances; well-known names and even signature phrases have been registered or protected as marks.
The IT Act and IT Rules, 2021. Impersonation content and synthetic media fall under intermediary obligations — platforms must act on valid complaints within defined timelines, which is what makes takedowns practical rather than theoretical.
3. How Indian courts have protected them
Over the past decade, Delhi and Bombay High Courts in particular have built a firm body of precedent. Amitabh Bachchan obtained a sweeping injunction in 2022 protecting his name, image and voice — including against misuse in lottery scams and unlicensed merchandise. Anil Kapoor won similar protection in 2023, with the Delhi High Court restraining unauthorised use of his image, voice and even his signature "jhakaas" delivery — notable because the order squarely addressed AI-generated content. In 2024, the Bombay High Court protected singer Arijit Singh against AI voice-cloning tools that let users generate songs in his voice, one of India's first orders aimed directly at generative-AI misuse.
The pattern across these cases: courts move fast when the identity is clearly established and the use is plainly commercial, and they increasingly grant technology-aware orders covering AI-generated content, domain names and future infringements — not just the specific instance complained about.
4. The DPDP Act layer
The Digital Personal Data Protection Act, 2023 adds a statutory front. Your face and voice are personal data; anyone processing them — including to train or run an AI model — needs a lawful basis, and consent under the Act must be specific, informed, and withdrawable. That last word matters most for talent: a blanket "you may use my content" from 2021 does not read as consent to voice-clone you in 2026. Rights infrastructure that records exactly what you consented to, for which use, for how long, converts DPDP from abstract compliance into a practical enforcement lever.
5. Generative AI changed the maths
Before generative AI, misusing a famous face took effort — a lookalike, a studio, a printer. Now a convincing face-swap takes minutes and a voice clone needs seconds of source audio, which is why unauthorised celebrity endorsements for betting apps and investment scams have become one of India's most common deepfake abuses. The law responded faster than many expected (see the cases above), but case-by-case litigation cannot match automated infringement. The practical answer is layered: register your identity so ownership is provable, monitor at machine speed, and enforce from a position of record. That layering is exactly what Zimorta's platform operationalises.
6. A practical protection playbook
Step 1 — Create the record. Register your likeness, voice and works with timestamps before you need them. Evidence assembled after an infringement is always weaker than a record that predates it. (Free during Zimorta early access.)
Step 2 — Make consent machine-readable. Decide your policy per use: GenAI training, AI ads, voice cloning, face swaps, political content. Written preferences that software can check beat verbal understandings every time. See machine-readable consent.
Step 3 — Price the licence, not the deliverable. Usage, territory, duration, exclusivity — each changes the number. Our free rate calculator shows the market-midpoint maths.
Step 4 — Monitor continuously. Most talent discover misuse from fans asking "is this really you?" — weeks late. Detection tooling finds deepfakes and unauthorised reuse while campaigns are still small.
Step 5 — Enforce in escalating order. Platform takedown first (fast, free), IT Rules notice next, legal notice citing your registration record third, litigation last. Each step is stronger and cheaper when steps 1–4 already happened.