Personality rights in India exist almost entirely because judges built them, not because Parliament legislated them, and 2026 has been the year that gap became impossible to ignore. In a single seven-month stretch, the Delhi and Bombay High Courts handed down a string of landmark orders protecting actors, a Member of Parliament, and public figures against AI chatbots, cloned voices, deepfake videos, and fabricated quotes, all without a single dedicated statute to point to. I run both a SaaS product and a media platform, so I sit on both sides of this issue, the technology that can generate this content, and the publisher obligations that follow when it spreads. In this guide, I’ll walk through exactly what Indian courts are protecting, how five back-to-back 2026 cases built a working legal framework out of constitutional principles and common law, the practical takedown protocol that’s emerged, and the four specific gaps Parliament still hasn’t addressed, gaps that matter enormously if you build AI products, run a platform, or are a public figure trying to protect your own identity.
Why India Still Has No Personality Rights Statute
Unlike copyright or trademark law, India has no dedicated Personality Rights Act. What exists instead is a patchwork built from three sources: Article 19(1)(a) of the Constitution (freedom of speech and expression, which courts have read to include control over one’s own persona), Article 21 (the right to life and personal dignity), and common law principles borrowed from the tort of passing off and provisions of the Copyright Act, 1957 (particularly moral rights under Section 57 and performer’s rights under Section 38). Courts have been building on this foundation since early cases like D.M. Entertainment v. Baby Gift House (2010) and ICC Development v. Arvee Enterprises, but 2026 is when this judge-made framework had to stretch to cover something entirely new: AI that doesn’t just copy an image, it generates a convincing new one, in your voice, saying things you never said.
The Five Cases That Defined AI-Era Personality Rights in 2026
Sonakshi Sinha v. Character Technologies Inc. & Ors. (March 2026) — The AI Chatbot Case
This was the first major case to target AI chatbot personas directly. Actor Sonakshi Sinha sued multiple defendants, including the US-based AI platform Character Technologies, alleging that unauthorised chatbots had been created using her name, image, and voice, generating lewd and obscene content and interacting with users as if she herself were responding. Justice Jyoti Singh of the Delhi High Court granted an ex parte ad interim injunction, restraining the use of Sinha’s “name, image, voice or likeness, directly or indirectly, for any purpose, including through AI tools such as chatbots, deepfakes or generative AI,” and ordered the identified infringing URLs taken down within 36 hours. The case is significant because it was the first to explicitly frame an AI chatbot persona, not just a static image or video, as a form of personality rights infringement.
Allu Arjun v. Ors. (April 2026) — Voice Cloning and “Fake Calls”
Telugu superstar Allu Arjun’s case, heard before Justice Tushar Rao Gedela of the Delhi High Court, expanded protection into genuinely new territory: AI voice-cloning tools that let users generate simulated phone calls in the actor’s voice, alongside unauthorised merchandise and sexually explicit deepfake content. The court’s order specifically named “Artificial Intelligence, Generative Artificial Intelligence, Machine Learning, Deepfakes, Face Morphing” as covered technologies, and recognised that Allu Arjun’s dialogues, mannerisms, and gestures, not just his face, had acquired protectable, distinctive status given his “iconic status” built through films like the Pushpa franchise.
Preity Zinta v. Google LLC & Ors. (June–July 2026) — The 275-URL Takedown and the Constitutional Basis
Preity Zinta’s case is arguably the most doctrinally important of the five. After securing leave from Justice Abhay Ahuja to sue Google and 16 other respondents (since several maintained offices outside Bombay’s jurisdiction), Justice Madhav Jamdar delivered an interim order in July 2026 directing Google, Meta, and X to take down or block roughly 275 infringing URLs within 72 hours. What makes this order significant is that Justice Jamdar explicitly grounded personality and publicity rights in Articles 19(1)(a) and 21 of the Constitution, and firmly rejected the idea that platforms could position themselves as passive conduits, stating they “cannot claim to be mere passive conduits” and are obligated to act. The court also directed an AI platform to add Zinta’s name to its character-creation and search blocklists, a forward-looking remedy aimed at preventing future infringement, not just cleaning up existing content.
Ravi Kishan v. Ashok Kumar (John Doe) & Ors. (July 2026) — Naming the Platforms Directly
Actor and sitting Member of Parliament Ravi Kishan’s case, also before Justice Jyoti Singh, is notable for explicitly naming Meta Platforms Inc., Google LLC, and X Corp. as entities obligated to act, directing them to disable access to offending content within 72 hours of notification if the original defendants failed to comply. The court reaffirmed that unauthorised use of a public figure’s identity affects not just commercial interests but the “right to lead a life of dignity,” reinforcing the Article 21 dignity-based reasoning that’s now a consistent thread across these orders.
Tabu v. Ashok Kumar & Ors. (August 2026) — Fabricated Quotes and Booking Fraud
Tabu’s case widened what counts as actionable misuse beyond deepfakes and chatbots entirely. Her suit alleged fake event-booking accounts (Instagram pages offering paid bookings for her appearances), fabricated statements falsely attributed to her and published online despite her categorical denial, unauthorised merchandise, and AI-generated obscene content. Justice Jyoti Singh ordered takedown of over 150 URLs, directed Google, Meta, X, and Reddit to disclose Basic Subscriber Information and IP logs to help identify the people behind the infringing accounts, and specifically restrained the publication circulating the fabricated statement, calling its continued circulation despite denial an “egregious act” apparently intended to “sensationalise the content and earn money.”
What’s Actually Protected Under Personality Rights in India (The Full List)
Reading across these five orders and the precedents they build on, Indian courts now recognise protection for:
- Name and stage name (including, as in Tabu’s case, both a performer’s real legal name and professional name)
- Image, photographs, and film stills
- Voice, including AI-cloned or synthetically generated voice
- Likeness and face, including deepfakes and face-morphed content
- Signature and initials
- Distinctive gestures, mannerisms, and dialogues (explicitly recognised in Allu Arjun’s case)
- Catchphrases (building on earlier precedent like Anil Kapoor’s protection of “jhakaas”)
- Fabricated quotes and statements falsely attributed to the person (new ground broken in Tabu’s case)
- Fake endorsements and booking fraud (also new in Tabu’s case)
This is a significantly broader list than most summaries of “personality rights” offer, and the expansion from image/voice into fabricated statements and booking fraud in just the last few months shows how quickly courts are stretching the doctrine to match new forms of misuse.
The Takedown Protocol Courts Have Developed
Across all five cases, a consistent procedural pattern has emerged, essentially a judge-made enforcement mechanism in the absence of statutory guidance:
- The suit is typically filed against named defendants plus “Ashok Kumar” (India’s equivalent of a John Doe order), covering unidentified infringers who can be added once identified.
- Courts increasingly require a clear, defendant-wise table of infringing URLs before granting relief — in Sonakshi Sinha’s case, the court specifically sent the plaintiff back to refile with clearer URL identification before passing the interim order, a procedural discipline worth knowing if you’re preparing your own suit.
- An ex parte ad interim injunction is granted, restraining use of the protected attributes across all named technologies, explicitly including AI, generative AI, deepfakes, and face morphing.
- A takedown window is fixed, typically 36 hours (Sonakshi Sinha) to 72 hours (Zinta, Ravi Kishan, broadly consistent with the statutory 3-hour SGI takedown window we’ve covered separately under the IT Amendment Rules 2026, though note these court-ordered windows are case-specific and distinct from that statutory timeline).
- Intermediaries are directed to disclose subscriber information, including Basic Subscriber Information and IP logs, to help identify anonymous infringers (seen clearly in Tabu’s order).
- Forward-looking remedies are increasingly included, such as blocklist directions preventing AI platforms from allowing new content to be generated using the plaintiff’s name or likeness, not just removing what already exists.
The Four Gaps Parliament Still Hasn’t Addressed
Despite this flurry of judicial activity, several structural problems remain that only legislation can fully fix:
1. No Standalone Statute Defining Personality or Publicity Rights
Every one of these five cases was decided using a combination of constitutional interpretation, common law passing-off principles, and Copyright Act provisions never designed for this purpose. This means protection depends entirely on how persuasively a plaintiff’s lawyers can assemble existing doctrine, rather than a clear statutory test, creating real unpredictability for both public figures and the platforms trying to comply.
2. No Statutory Framework for Post-Mortem (Deceased) Personality Rights
Indian courts have given mixed signals on whether personality rights survive death and for how long, unlike jurisdictions such as several US states that have clear statutory post-mortem publicity rights terms. For legacy estates, family members of deceased public figures, and AI tools that can now recreate a deceased performer’s voice or likeness with unsettling accuracy, this is an urgent, unaddressed gap.
3. No AI-Developer-Specific Liability Framework
The current approach treats AI chatbot platforms, voice-cloning tool providers, and deepfake generators as intermediaries under the general IT Rules framework, rather than creating purpose-built obligations for companies whose core product function is generating synthetic human likeness. This matters enormously if you’re building in this space, since the compliance bar you’re held to is currently assembled from court orders case by case, rather than a clear statutory standard you can design against from day one.
4. No Clear Statutory Definition of “Voice” as a Protectable Biometric Attribute
While courts have consistently included voice within the attributes protected (Allu Arjun’s voice-cloning claim, Tabu’s “voice recordings and duplications”), there’s no statutory definition establishing voice as a distinct, protectable biometric identity marker the way some global frameworks are beginning to do, leaving its exact scope to continued judicial interpretation.
What This Means for Brands, Tech Companies, and AI Developers
If you’re building or running a platform that touches any of this, the practical takeaway is sobering: courts are not waiting for Parliament, and neither should your compliance posture. A few specific implications:
- If your SaaS product includes voice-cloning, face-swapping, or chatbot-persona features, you are now operating in an area where Indian courts have shown willingness to issue broad, fast-moving injunctions and blocklist directions, even against foreign-incorporated platforms, as Character Technologies’ case shows.
- If you run a platform hosting user-generated content, the “passive conduit” defence has been explicitly rejected in Preity Zinta’s case, and this sits directly alongside your separate obligations under the IT Amendment Rules 2026’s Synthetically Generated Information framework, meaning a single piece of infringing content can trigger both personality rights liability and SGI takedown obligations simultaneously.
- If you’re a brand considering AI-generated endorsements or influencer content, using any recognisable public figure’s likeness, voice, or mannerisms without explicit authorisation, even through an AI tool rather than a direct photo, now carries real injunction risk backed by five separate 2026 precedents.
- If you’re advising or representing a public figure, the procedural lesson from Sonakshi Sinha’s case is clear: have your defendant-wise, URL-specific evidence organised before filing, since courts are increasingly strict about this at the interim relief stage.
If you’re building a product in this space or need to assess your platform’s exposure under this emerging framework, it’s worth getting a specific compliance review of your AI features against current personality rights precedent, since this is an area where the law is moving faster than most generic legal guidance can keep up with.
How Personality Rights Differ From Related Legal Concepts
| Personality Rights | Copyright | Trademark | |
|---|---|---|---|
| What’s protected | Name, image, voice, likeness, mannerisms | Original creative works (films, music, writing) | Brand names, logos, commercial marks |
| Statutory basis in India | None dedicated; built from Articles 19(1)(a), 21, common law, Copyright Act | Copyright Act, 1957 | Trade Marks Act, 1999 |
| Who can claim it | Any individual, especially public figures with commercial goodwill | Original creator or assignee | Registered mark owner |
| Survives death? | Unclear, no statutory answer | Yes, for a defined statutory term | Yes, if renewed |
| 2026 AI relevance | Central battleground (all five cases above) | Relevant for AI training data disputes | Less directly relevant |
Frequently Asked Questions About Personality Rights in India
1. Is there a specific law for personality rights in India?
No. India has no standalone personality rights or publicity rights statute. Protection is built through judicial interpretation of Articles 19(1)(a) and 21 of the Constitution, the common law tort of passing off, and provisions of the Copyright Act, 1957, as seen across the Sonakshi Sinha, Allu Arjun, Preity Zinta, Ravi Kishan, and Tabu cases of 2026.
2. What exactly counts as a violation of personality rights?
Based on recent precedent, this includes unauthorised use of a person’s name, image, voice, signature, distinctive gestures or mannerisms, catchphrases, and now, following the Tabu case, fabricated quotes falsely attributed to someone and fake booking or endorsement accounts.
3. Can AI chatbots impersonating a celebrity be legally stopped in India?
Yes. The Sonakshi Sinha v. Character Technologies case specifically established that AI chatbot personas built using a person’s name, image, and voice without authorisation constitute a violation of personality rights, with courts ordering takedown of infringing platforms and content.
4. How fast must platforms take down infringing AI content after a court order?
Timelines have ranged from 36 hours (Sonakshi Sinha’s case) to 72 hours (Preity Zinta, Ravi Kishan), set case by case by the court, which is distinct from the separate 3-hour statutory takedown window for Synthetically Generated Information under the IT Amendment Rules 2026.
5. Do personality rights only apply to Bollywood actors and celebrities?
No, though most reported cases involve public figures given their higher commercial stakes and visibility. The underlying legal principles, dignity under Article 21 and control over commercial exploitation of one’s identity, apply to any individual, though proving commercial harm or distinctiveness may be harder for private individuals.
6. Can voice cloning alone, without any image misuse, be legally actionable?
Yes. The Allu Arjun case specifically addressed AI voice-cloning tools used to simulate “fake calls” in the actor’s voice, and the court’s injunction explicitly covered voice cloning as a distinct form of personality rights infringement, separate from image-based deepfakes.
7. What is a “John Doe” or “Ashok Kumar” order in these cases?
It’s an order filed against unidentified or anonymous defendants (referred to as “Ashok Kumar” in Indian practice, the equivalent of “John Doe”), allowing plaintiffs to act against infringers whose identity isn’t yet known, with intermediaries later directed to disclose subscriber information to help identify them.
8. Do these personality rights protections extend to deceased celebrities?
This remains legally unclear in India. Unlike some foreign jurisdictions with defined statutory post-mortem publicity rights, India has no clear statutory answer on whether, or for how long, personality rights survive after death, making this one of the most significant gaps Parliament has yet to address.
9. Can a platform be held liable even if it didn’t create the infringing AI content itself?
Yes, if it fails to act on takedown requests after being notified. The Preity Zinta case specifically rejected the argument that platforms like Google and Meta are merely “passive conduits,” holding they have an active obligation to act against infringing content once flagged.
10. How does this relate to India’s IT Amendment Rules 2026 on AI content?
The two frameworks operate in parallel. Personality rights cases proceed through civil suits seeking injunctions and damages for the individual harmed, while the IT Amendment Rules 2026 impose separate, statutory due diligence and takedown obligations on platforms handling Synthetically Generated Information generally, meaning a single piece of AI-generated content misusing someone’s identity can trigger obligations under both frameworks simultaneously.
Conclusion
Five cases in seven months is not a coincidence, it’s a signal that Indian courts have decided they cannot wait for Parliament to legislate a personality rights framework while AI-generated chatbots, cloned voices, and fabricated statements spread faster than any statute could be drafted. What’s emerged instead is a genuinely workable, if judge-made, enforcement system: constitutional grounding under Articles 19(1)(a) and 21, a consistent takedown protocol, and an expanding list of protected attributes that now reaches well beyond simple image misuse into voice cloning, chatbot personas, and fabricated quotes. But the four gaps that remain, no dedicated statute, no post-mortem rights clarity, no AI-developer-specific liability framework, and no statutory definition of voice as a protectable attribute, mean real uncertainty persists for anyone building AI products, running a platform, or trying to protect their own identity in this space.
If you’re a public figure dealing with AI misuse of your identity, or a founder trying to understand how these precedents affect your product’s compliance obligations, it’s worth getting a specific assessment against this fast-moving body of case law, rather than relying on a generic content policy that hasn’t caught up to where courts already are.