Celebrity Personality Rights in India: Can AI Use Your Face or Voice Without Consent?

Introduction

Imagine opening Instagram and discovering an advertisement in which you appear to recommend a financial product you have never used.

The face is yours.

The voice sounds exactly like yours.

The mannerisms are recognisable.

But you never recorded the advertisement.

Artificial intelligence generated it.

For celebrities, actors, singers, sportspersons, influencers and other public figures, this is no longer a hypothetical problem.

Artificial intelligence can now reproduce a person’s:

• face;
• voice;
• manner of speaking;
• singing style;
• gestures;
• facial expressions;
• likeness;
• digital avatar; and
• other recognisable personality attributes

with extraordinary accuracy.

Indian courts have consequently been confronted with a new generation of disputes involving AI-generated celebrity voices, deepfake videos, face-swapped images, synthetic endorsements, fake merchandise, AI chatbots and digital avatars.

The central legal question is:

Can an AI company, advertiser, social-media creator or other person use a celebrity’s face or voice without consent?

The short answer is:

Not automatically.

Where a person’s identity is recognisable and their face, voice, likeness or other personality attributes are used without authorisation—particularly for commercial exploitation, fake endorsement, merchandising, impersonation or misleading synthetic content—Indian courts have increasingly recognised remedies based on personality rights, publicity rights, privacy, passing off, trademark law, copyright-related rights and other legal principles.

Recent judgments have gone even further.

In 2026, Indian courts have expressly restrained defendants from using celebrities’ personalities through technologies including:

• Artificial Intelligence;
• Generative AI;
• deepfakes;
• AI chatbots;
• face swapping;
• voice cloning;
• synthetic media; and
• digital avatars.

However, personality rights are not unlimited.

A celebrity cannot necessarily prohibit every photograph, meme, parody, news report, criticism, artistic reference or public discussion involving them.

Indian law must balance:

personality and privacy rights

against

freedom of speech, artistic expression, news reporting, parody and legitimate public interest.

This article examines where Indian law currently draws that line.

What Are Personality Rights?

Personality rights broadly refer to an individual’s ability to control the unauthorised exploitation of attributes that identify that person.

These may include:

• name;
• photograph;
• image;
• likeness;
• voice;
• signature;
• distinctive appearance;
• catchphrases;
• mannerisms;
• style of speaking;
• manner of singing;
• persona; and
• other recognisable attributes.

Personality rights are often discussed together with publicity rights, although the concepts overlap rather than being perfectly identical.

A publicity right is especially concerned with the commercial value of identity.

For example, if a famous actor ordinarily charges ₹5 crore to endorse a luxury brand, an unrelated business should not ordinarily be permitted to generate an AI advertisement using the actor’s face and voice in order to create the false impression that the celebrity endorsed its product.

The unauthorised advertiser would effectively be appropriating the commercial value of the celebrity’s identity.

Does India Have a Separate “Personality Rights Act”?

No.

India does not presently have one comprehensive standalone statute titled the “Personality Rights Act” or “Right of Publicity Act”.

Instead, personality rights have developed through a combination of:

• constitutional privacy principles;
• common-law passing off;
• intellectual-property law;
• trademark protection;
• copyright and performer’s rights;
• tort principles;
• unfair commercial exploitation; and
• judicial precedent.

The Delhi High Court has played a particularly important role in developing this doctrine.

In ICC Development (International) Ltd. v. Arvee Enterprises, the Delhi High Court explained that publicity rights arise from privacy and belong to individuals. The Court specifically recognised that personality indicia can include matters such as a person’s name, personality traits, signature and voice.

The jurisprudence has developed significantly since then.

Personality Rights vs Privacy Rights

The two concepts are related but distinct.

Privacy Rights

Privacy protects an individual’s autonomy, dignity and personal sphere from unjustified intrusion.

The constitutional foundation lies primarily in Article 21 of the Constitution, reinforced by the Supreme Court’s landmark decision in Justice K.S. Puttaswamy (Retd.) v. Union of India.

Privacy concerns may arise where AI is used to create:

• fake intimate videos;
• sexually explicit deepfakes;
• fabricated private conversations;
• false representations;
• manipulated personal imagery; or
• other intrusive synthetic content.

Personality or Publicity Rights

Publicity rights primarily address control over the commercial exploitation of identity.

The question is often:

Has someone commercially appropriated the person’s identity without permission?

Accordingly, the same AI deepfake can potentially affect both:

privacy, because it manipulates personal identity;

and

publicity rights, because it commercially exploits that identity.

Can a Celebrity Own Their Face?

Not in the same way that someone owns a physical object.

The law does not mean that a celebrity has absolute ownership over every photograph, reference or representation of their face.

Instead, courts protect the individual’s legally recognised interest in preventing unauthorised exploitation of their identity in certain circumstances.

The critical questions include:

• Is the person identifiable?
• Is their identity being commercially exploited?
• Does the use falsely imply endorsement?
• Is the content misleading?
• Does it damage reputation or dignity?
• Is it parody, criticism, commentary or legitimate news?
• Has consent been obtained?
• Does another intellectual-property right apply?

Anil Kapoor v. Simply Life India: AI Enters Personality Rights Litigation

One of India’s most important decisions concerning AI and celebrity identity is Anil Kapoor v. Simply Life India & Ors., decided by the Delhi High Court in September 2023.

Anil Kapoor sought protection against unauthorised commercial exploitation of attributes including:

• his name;
• image;
• likeness;
• voice;
• signature;
• manner of speaking; and
• other features associated with his personality.

The Delhi High Court granted an interim injunction.

Crucially, the Court expressly restrained defendants from misusing Kapoor’s personality through technological tools including:

Artificial Intelligence, machine learning, deepfakes and face morphing.

The Court found a prima facie violation of personality rights and privacy, noting the risk of commercial exploitation and harm to dignity.

This decision became an important precedent for subsequent AI-personality-rights litigation.

Arijit Singh: AI Voice Cloning and Singing Style

The Bombay High Court’s 2024 ruling concerning singer Arijit Singh is particularly important for generative AI.

The case concerned unauthorised exploitation of:

• his name;
• voice;
• vocal style;
• vocal techniques;
• manner of singing;
• photographs;
• likeness; and
• other elements of his personality.

The Court restrained defendants from using AI technologies to imitate Singh’s voice and persona.

Most significantly, the Bombay High Court observed that making AI tools available which enable users to convert voices into that of a celebrity without permission can violate the celebrity’s personality rights.

The injunction covered technologies including:

• AI voice models;
• voice-conversion tools;
• synthesised voices;
• digital avatars;
• generative AI;
• machine learning;
• deepfakes; and
• face morphing.

The judgment is particularly significant for singers.

Their commercially valuable identity may lie not merely in their photograph but in the distinctive qualities of their voice and performance.

Can AI Clone a Singer’s Voice Without Consent?

The safer legal answer in India today is:

Commercial voice cloning of an identifiable celebrity without authorisation can expose the creator or business to substantial legal risk.

The Arijit Singh decision strongly supports this conclusion.

In February 2026, the Delhi High Court also granted protection to singer Jubin Nautiyal, including against the creation or use of:

• artificial-intelligence voice models;
• voice-conversion tools;
• synthesised voices;
• digital avatars; and
• technology imitating personality attributes.

The direction of Indian jurisprudence is therefore increasingly clear.

A person’s distinctive voice can be an important protectable element of their persona.

What About AI Mimicking Singing Style?

This becomes more complex.

Copyright law traditionally protects original works and specified performances; it does not simply grant ownership over every abstract artistic “style”.

However, personality-rights litigation can extend beyond conventional copyright infringement where an AI output deliberately imitates an identifiable singer so closely that the public recognises the synthetic performance as being that person.

In Arijit Singh, the Bombay High Court’s protection expressly referred to aspects including his vocal style, technique, arrangements and manner of singing.

Thus, an argument such as—

“We did not copy the original recording; we only made an AI voice that sounds exactly like the singer”

does not necessarily eliminate liability.

Performer’s Rights Under the Copyright Act

The Copyright Act, 1957 separately recognises statutory rights of performers.

Section 38 gives a performer a special performer’s right.

Section 38A provides exclusive rights concerning recording, reproduction, distribution and communication of a performance.

Section 38B also grants performers certain moral rights, including rights concerning identification and prejudicial distortion or modification of their performances.

These statutory rights can overlap with personality-rights disputes.

However, it is important to distinguish:

copying an existing performance

from

generating a new synthetic performance imitating the performer.

The latter may raise personality/publicity issues even where conventional copyright analysis becomes more complicated.

The Explosion of Personality-Rights Cases in 2025–2026

Indian courts have seen a remarkable rise in cases seeking protection against AI-generated exploitation of personality.

Recent plaintiffs have included actors, singers, athletes, filmmakers and public figures.

The cases illustrate how quickly personality-rights doctrine is expanding.

Karan Johar and the Limits of Personality Rights

In Karan Johar v. Ashok Kumar/John Doe & Ors., the Delhi High Court in September 2025 restrained unauthorised commercial use of Karan Johar’s:

• name;
• “KJo” identity;
• likeness;
• image;
• voice; and
• other personality attributes,

including misuse through AI, machine learning, deepfakes, face morphing and similar technologies.

However, the Court also declined at that stage to order removal of certain URLs that appeared to fall within recognised concepts such as parody, caricature and lampooning.

This is extremely important.

Personality rights do not automatically create a censorship right over every reference to a celebrity.

Abhishek Bachchan: AI, Misleading Context and Privacy

In September 2025, the Delhi High Court granted interim protection to Abhishek Bachchan against unauthorised use of his name, image, voice and personality.

The Court noted that technologically generated representations placing him in misleading, inappropriate or derogatory settings could interfere with both personality rights and privacy.

This illustrates how AI deepfakes may engage several legal interests at the same time:

• commercial personality rights;
• privacy;
• reputation;
• dignity; and
• potential defamation.

Gautam Gambhir: False Endorsement and Merchandise

In March 2026, the Delhi High Court dealt with unauthorised use of Gautam Gambhir’s personality attributes.

The allegations included merchandise sold through e-commerce platforms using his image in a way that allegedly suggested affiliation or endorsement, alongside morphed and AI-generated content.

The Court reiterated that unauthorised commercial use of a person’s name, image and distinctive attributes can dilute their identity and produce unearned commercial gain for others.

False endorsement is therefore one of the clearest situations in which personality rights may become enforceable.

Sonakshi Sinha and AI Chatbots

In March 2026, the Delhi High Court granted interim protection to Sonakshi Sinha against unauthorised use of:

• her name;
• image;
• voice;
• likeness; and
• other personality attributes

through technologies including:

• AI;
• Generative AI;
• machine learning;
• deepfakes;
• AI chatbots; and
• face morphing.

The Court restrained specified defendants from exploiting these attributes without consent for commercial or personal gain and directed removal of identified URLs.

The reference to AI chatbots is significant.

A chatbot designed to impersonate a real celebrity may create personality-rights issues even where no conventional video deepfake exists.

Varun Dhawan: Commercial Exploitation Through Generative AI

In May 2026, the Delhi High Court protected actor Varun Dhawan’s name, image, voice and likeness against unauthorised commercial exploitation.

The injunction expressly covered:

• Artificial Intelligence;
• Generative AI;
• machine learning;
• deepfakes;
• AI chatbots; and
• face morphing.

The Court also restrained unauthorised merchandise and misleading use of personality attributes.

Ravi Kishan and Synthetic Personality Exploitation

In July 2026, the Delhi High Court restrained defendants from exploiting Ravi Kishan’s personality attributes through AI, Generative AI, deepfakes and other technologies.

The protection extended to his:

• name;
• image;
• likeness; and
• other recognisable personality attributes.

Yuvraj Singh: Deepfakes, Face Swapping and Speech Style

The Delhi High Court’s July 2026 order concerning Yuvraj Singh is another important development.

The Court protected:

• “Yuvraj Singh”;
• “Yuvi”;
• his image;
• voice;
• likeness; and
• other personality attributes.

The injunction specifically addressed AI-generated representations, deepfakes, face swapping, morphing, superimposing and even use of recognisable style of speech.

This demonstrates that personality-rights protection may extend beyond a literal photograph or sound recording.

The inquiry increasingly focuses on whether the synthetic output appropriates an individual’s recognisable identity.

Khushi Kapoor: Voice Cloning and Synthetic Media

In August 2026, the Delhi High Court granted interim protection to actor Khushi Kapoor.

The Court recognised protectable interests in her:

• name;
• image;
• likeness;
• voice; and
• other personality attributes.

The injunction expressly extended to:

• AI;
• Generative AI;
• deepfakes;
• AI chatbots;
• face morphing;
• voice cloning; and
• synthetic media.

This is one of the clearest contemporary judicial responses to generative-AI identity exploitation.

Do You Need to Be a Celebrity to Have Rights Over Your Identity?

Not necessarily.

Celebrity personality-rights cases are particularly visible because celebrities have:

• significant goodwill;
• recognisable personas;
• valuable endorsement markets; and
• measurable commercial value attached to identity.

But privacy, identity, impersonation and data-protection principles are not restricted only to famous people.

An ordinary person whose face or voice is cloned for:

• fraud;
• pornography;
• fake advertisements;
• financial scams;
• impersonation; or
• other unlawful purposes

may potentially have remedies under privacy law, data-protection law, the Information Technology Act, criminal law and other doctrines depending upon the facts.

The commercial “publicity right” analysis may be strongest for individuals whose persona carries established commercial value, but the underlying rights to privacy and identity are not exclusively celebrity rights.

Digital Personal Data Protection Act and AI Identity

The Digital Personal Data Protection Act, 2023, together with the DPDP Rules, 2025, now forms India’s general digital-personal-data framework.

The Government notified the DPDP Rules in November 2025, bringing the framework into operation.

A person’s identifiable digital photograph, facial information or voice recording may constitute digital personal data where it relates to an identifiable individual.

Accordingly, collection or processing of such material for AI systems can potentially raise DPDP questions concerning:

• lawful processing;
• consent;
• notice;
• purpose;
• data minimisation;
• security; and
• rights of the Data Principal.

However, DPDP liability must be analysed separately from personality-rights liability because the statute contains its own scope, grounds of processing and exceptions.

A personality-rights infringement does not automatically mean that every DPDP provision has necessarily been violated, and vice versa.

Deepfakes and the Information Technology Rules

India’s regulatory response to synthetic media has also changed significantly.

On 10 February 2026, the Government amended the Information Technology Rules, 2021 to strengthen regulation of synthetically generated information, including deepfakes and AI-generated audio-visual content.

According to the Ministry of Electronics and Information Technology, the amended framework includes requirements concerning:

• identification and labelling of synthetic content;
• traceable metadata;
• intermediary due diligence;
• user accountability;
• impersonation;
• unlawful deepfakes;
• stronger grievance mechanisms; and
• faster removal processes for certain unlawful content.

The Government has specifically identified deepfake impersonation as a serious regulatory concern.

Information Technology Act and Deepfake Impersonation

Depending upon what an AI-generated impersonation is used for, the Information Technology Act may also become relevant.

The Government has identified provisions including:

• Section 66C — identity theft;
• Section 66D — cheating by personation using computer resources;
• Section 66E — violation of privacy;
• Section 67 — obscene electronic material; and
• Section 67A — sexually explicit electronic material

as potentially relevant to deepfake harms.

For example, an AI clone of a celebrity’s voice used to convince victims to invest money could involve issues very different from an AI parody song.

The purpose and context matter.

Can AI Use a Celebrity’s Face in an Advertisement Without Permission?

This is one of the clearest scenarios.

Suppose a company generates a video showing a well-known actor apparently saying:

“I use this investment app. You should download it too.”

The actor never consented.

This could potentially involve:

• violation of personality/publicity rights;
• false endorsement;
• passing off;
• misleading advertising;
• privacy concerns;
• trademark issues;
• consumer-protection issues; and
• potentially fraud if used deceptively.

The commercial value of celebrity identity exists largely because consumers may change purchasing decisions based on perceived endorsement.

An advertiser should not ordinarily be permitted to obtain that value for free by generating the endorsement artificially.

What Is Passing Off?

Passing off is a common-law remedy traditionally associated with protecting business goodwill against misrepresentation.

In personality-rights cases, the doctrine can become relevant where unauthorised use of a celebrity’s identity creates a false suggestion that:

• the celebrity endorsed a product;
• approved a service;
• is commercially associated with a brand; or
• authorised merchandise.

For example:

“Official Virat Kohli Fitness Supplement”

would create serious legal issues if Virat Kohli had no involvement with the product.

The wrongdoing is not merely copying a photograph.

It is misleading consumers into believing there is an endorsement or commercial connection.

Trademark Law and Celebrity Names

Celebrities increasingly register:

• their names;
• stage names;
• initials;
• signatures;
• logos;
• catchphrases; or
• other commercial identifiers

as trademarks.

Trademark registration can provide an additional cause of action against unauthorised commercial exploitation.

However, personality rights do not depend entirely on trademark registration.

Indian courts have protected personality attributes even where the central claim rests upon publicity, privacy or passing-off principles rather than exclusively on a registered trademark.

Copyright and Celebrity Photographs

An important distinction must be made between:

the celebrity appearing in a photograph

and

copyright in the photograph itself.

A celebrity does not automatically own copyright in every photograph taken of them.

Copyright ownership may belong to a photographer, employer, producer or other rights holder depending upon the circumstances.

But a business that lawfully acquires a licence to use a photograph may still face a separate personality-rights problem if it uses the image to falsely imply that the celebrity endorses a product.

Thus:

copyright permission and personality permission are not necessarily the same thing.

Example

Suppose a stock-photo agency legally licences an image of a famous cricketer to a business.

The business then places the photograph next to the statement:

“Recommended by India’s leading cricketers.”

Even if the photograph licence is valid, the advertisement could potentially create an unauthorised endorsement implication.

The legal issue therefore goes beyond copyright.

Can AI Train on Celebrity Images or Voices?

This is considerably more complicated than using the resulting clone in an advertisement.

Training generative-AI systems raises multiple questions concerning:

• copyright;
• data protection;
• privacy;
• contract;
• database access;
• personality rights; and
• output generation.

Indian law has not yet produced a single Supreme Court judgment comprehensively determining whether all training of AI models on publicly accessible celebrity photographs or voices requires personality-rights consent.

The strongest Indian decisions so far generally concern outputs and exploitation—for example, AI-generated voices, deepfake videos, synthetic merchandise and commercial impersonation.

Therefore, statements such as—

“Training on public celebrity images is always legal”

or

“Every AI company must obtain permission before any model training”

would presently be too absolute.

This area remains legally developing.

Can AI Generate a Celebrity-Like Voice Without Using Their Name?

Potentially, yes—but legal risk remains if the person is readily identifiable.

Personality rights focus heavily on identifiability.

A company cannot necessarily avoid liability simply by saying:

“We never used the celebrity’s name.”

If ordinary listeners immediately recognise the synthetic voice as imitating a specific singer or actor, courts may consider whether the person’s persona has effectively been appropriated.

The Arijit Singh litigation demonstrates how voice, singing mannerisms and associated traits may collectively identify a performer.

What If the AI Disclaimer Says “Not Really the Celebrity”?

A disclaimer may help reduce deception in some contexts, but it does not automatically legalise commercial exploitation.

Suppose a company uses an unmistakable AI recreation of a superstar in an advertisement and writes in tiny letters:

“AI generated. Celebrity not actually affiliated.”

The court may still examine whether the advertiser appropriated the commercial value of the celebrity’s identity without consent.

Disclosure and consent are different concepts.

A label can tell viewers something is synthetic.

It does not necessarily grant the creator the right to exploit another person’s identity commercially.

What About AI Fan Art?

The answer depends heavily upon context.

A non-commercial fan image created for personal expression may present a very different legal case from:

• selling merchandise;
• operating a paid AI service;
• monetising advertising;
• creating a fake brand endorsement;
• selling NFTs;
• using the personality to drive subscriptions; or
• misleading viewers.

Commercial exploitation generally strengthens a personality-rights claim.

What About Memes?

A celebrity does not own every meme that includes their face.

Memes may involve:

• humour;
• satire;
• parody;
• commentary;
• criticism; and
• cultural expression.

Courts must balance such speech against personality and privacy rights.

However, the context can shift where the meme:

• falsely appears to be an advertisement;
• is sold commercially;
• is defamatory;
• is sexually explicit;
• involves manipulated intimate content; or
• otherwise crosses recognised legal boundaries.

Parody and Satire

The Karan Johar proceedings illustrate an important limitation.

The Delhi High Court recognised that some content could potentially fall within established protections relating to parody, caricature and lampooning, and therefore did not indiscriminately order every impugned URL removed.

This reflects the constitutional importance of Article 19(1)(a).

Personality rights should not become a mechanism through which public figures suppress every joke or critical reference.

Can News Organisations Use Celebrity Images Without Consent?

Generally, legitimate news reporting and matters of public interest are treated differently from commercial endorsement.

If a newspaper publishes a photograph of an actor while reporting:

“Actor wins National Film Award”

that is conceptually different from a company using the same image to say:

“Actor recommends our insurance policy.”

Personality rights do not ordinarily amount to a veto over legitimate journalism.

However, newsworthiness is also not a blank cheque for fabricated or misleading AI content.

A news organisation publishing a realistic AI-generated video falsely showing a celebrity committing misconduct could face entirely different legal issues.

AI Deepfake Pornography

Non-consensual sexually explicit deepfakes represent one of the most serious forms of identity misuse.

Such content may engage:

• privacy rights;
• dignity;
• defamation;
• IT Act offences;
• obscenity and sexually explicit-content provisions;
• intermediary-removal mechanisms; and
• potentially criminal law depending upon the facts.

The 2026 IT-rule framework specifically strengthens platform obligations concerning unlawful synthetic content and sensitive categories including impersonation and non-consensual intimate imagery.

Commercial personality-rights doctrine is therefore only one part of the available legal response.

AI Fraud Using Celebrity Faces

Fraudsters increasingly use deepfake videos of public figures to promote:

• cryptocurrency schemes;
• fake investment platforms;
• betting websites;
• miracle health products;
• financial scams; and
• fraudulent mobile applications.

Such content can create multiple simultaneous causes of action.

For the celebrity:

• personality-rights infringement;
• reputational injury;
• false endorsement.

For consumers:

• fraud;
• deception;
• misleading advertisement.

For law enforcement:

• cybercrime;
• personation;
• potentially financial offences.

AI deepfakes therefore affect both the person whose identity is stolen and the public who may be deceived by it.

Can an AI Company Create a Celebrity Chatbot?

This has become an increasingly significant question.

Suppose an AI service advertises:

“Talk privately with your favourite Bollywood star.”

Users then communicate with an AI character deliberately designed to replicate a real actor’s:

• name;
• photograph;
• voice;
• conversational style; and
• personality.

If this is done without authorisation and monetised through subscriptions, the service could face a substantial personality-rights claim.

The Sonakshi Sinha and several 2026 Delhi High Court orders expressly refer to AI chatbots in personality-rights injunctions.

What About Dead Celebrities?

Post-mortem personality rights in India remain more legally complex and less settled than rights of living celebrities.

Questions may involve:

• estates;
• heirs;
• trademark ownership;
• copyright;
• contractual rights;
• merchandising rights; and
• passing off.

Indian law does not yet provide a simple nationwide statutory rule stating that personality rights automatically continue for a fixed number of years after death in every circumstance.

Businesses therefore should not assume that a deceased celebrity’s identity is automatically free for commercial exploitation.

Existing intellectual-property rights and estate interests may remain highly relevant.

Can a Celebrity Sell or License Personality Rights?

Commercial use can certainly be licensed.

Celebrity endorsement contracts routinely permit authorised use of:

• name;
• image;
• likeness;
• signature;
• photographs;
• voice; and
• approved content

subject to contractual restrictions.

Modern contracts increasingly need to address AI specifically.

For example:

Does permission to use a recorded voice also permit training an AI voice clone?

Can the brand create new synthetic advertisements after the celebrity stops recording?

Can the AI-generated likeness be used after the contract expires?

Traditional endorsement language may not adequately answer these questions.

AI Clauses Are Becoming Essential in Celebrity Contracts

A modern endorsement or entertainment agreement should consider expressly addressing:

• whether AI training is permitted;
• whether synthetic replicas can be created;
• voice cloning;
• face cloning;
• digital avatars;
• permitted prompts and outputs;
• approval rights;
• duration;
• territory;
• advertising channels;
• data retention;
• deletion of AI models;
• use after termination;
• sublicensing;
• security;
• third-party model providers; and
• compensation for synthetic reuse.

Without such language, disputes may arise over whether an old-fashioned image licence implicitly allowed AI-generated derivative exploitation.

How Can a Celebrity Enforce Personality Rights?

A person facing unauthorised use may seek civil remedies.

Depending upon the facts, these can include:

• interim injunction;
• permanent injunction;
• removal of infringing material;
• disclosure of infringers’ identities;
• blocking of websites;
• suspension of domain names;
• damages;
• accounts of profits;
• delivery up of infringing merchandise; and
• other appropriate relief.

Indian celebrity cases increasingly use John Doe/Ashok Kumar defendants where the identities of online infringers are initially unknown.

Courts may also direct social-media intermediaries, marketplaces, registrars or government authorities to assist in disabling identified content.

For example, the 2026 Khushi Kapoor order included directions concerning takedown of numerous URLs.

Why Are John Doe Orders Important?

AI-generated infringing content can spread anonymously and extremely rapidly.

A deepfake may be reposted by hundreds of unknown accounts before a celebrity can identify the original creator.

John Doe orders allow courts to grant relief against unidentified infringers described by placeholder names such as:

“Ashok Kumar” or “John Doe.”

This makes it possible to obtain urgent relief without waiting until every anonymous creator has been identified.

Can Platforms Be Ordered to Remove AI Deepfakes?

Yes, subject to applicable law and court directions.

Courts in recent personality-rights cases have directed intermediaries and platforms to disable access to identified infringing URLs.

India’s amended IT Rules also impose strengthened obligations concerning unlawful synthetically generated content.

However, platform liability and intermediary safe-harbour protection remain separate legal questions that depend upon compliance with the Information Technology Act and rules.

How Quickly Can Harm Occur?

AI substantially changes the nature of personality-rights litigation.

Traditionally, unauthorised celebrity merchandise might involve hundreds of physical T-shirts.

A generative-AI model can now create thousands of personalised advertisements in minutes.

A single voice-cloning tool can allow millions of users to make a famous singer appear to sing words that the singer never performed.

A deepfake can become viral internationally within hours.

This is why courts frequently grant urgent interim or ex parte injunctions in personality-rights cases.

By the time ordinary litigation reaches final judgment, the damage may already be impossible to reverse.

Freedom of Speech vs Personality Rights

This is likely to become one of the most important issues in future Indian AI litigation.

Personality-rights protection is necessary to prevent exploitation.

But overly broad rights could interfere with:

• political cartoons;
• film criticism;
• satire;
• parody;
• biographies;
• documentaries;
• fan culture;
• journalism;
• academic analysis; and
• creative expression.

Indian law therefore cannot simply adopt the rule:

“Any unauthorised use of a celebrity’s face is illegal.”

The context matters.

The Bombay High Court itself recognised in the Arijit Singh litigation that freedom of speech permits critique and commentary, while holding that it does not provide unrestricted freedom to commercially exploit a celebrity’s persona.

Public Figure Does Not Mean Public Property

Celebrities voluntarily enter public life.

That means they inevitably receive less practical privacy concerning matters genuinely in the public domain than an ordinary private citizen.

But becoming famous does not mean surrendering all control over one’s identity.

A person’s fame is often precisely what gives their personality commercial value.

If public visibility automatically destroyed the right to control commercial identity, anyone could use:

• Shah Rukh Khan to advertise a housing project;
• a famous cricketer to promote a betting app;
• a singer’s cloned voice to sell music software;
• an actor’s deepfake to promote cryptocurrency,

without paying for endorsement rights.

The personality-rights doctrine exists in part to prevent that unauthorised appropriation.

Is Commercial Use Necessary for Every Personality Claim?

Commercial exploitation is a central element of many publicity-rights cases.

Several Indian judgments emphasise unauthorised commercial gain.

However, modern AI cases also involve:

• privacy;
• dignity;
• defamation;
• obscenity;
• malicious impersonation; and
• reputational harm.

Therefore, a person may potentially have legal remedies even where the infringer did not directly sell a product.

For example, an obscene deepfake generated solely to humiliate an actress may raise serious legal claims even without conventional commercial exploitation.

The correct cause of action depends upon the facts.

What Does the Plaintiff Normally Need to Show?

Although the precise requirements depend upon the cause of action, relevant factors often include:

1. Identifiability

The disputed content must relate recognisably to the claimant.

2. Personality value or goodwill

For classic publicity-rights claims, particularly celebrity cases, courts examine reputation and commercial recognition.

3. Unauthorised use

The defendant lacked valid consent or licence.

4. Commercial exploitation or legally cognisable harm

The use may generate profit, imply endorsement, dilute identity, infringe privacy or cause reputational injury.

The Bombay High Court in Arijit Singh emphasised celebrity status, identifiability and unauthorised exploitation for commercial or personal gain.

Can a Celebrity Trademark Their Voice?

Trademark protection for non-traditional marks is legally more complicated than simply registering a name.

But celebrity names, signatures, catchphrases and other distinctive indicia may sometimes receive trademark protection where statutory requirements are satisfied.

Voice-based trademarks raise additional questions concerning graphical representation, distinctiveness and source identification.

Therefore, personality-rights protection may remain more practical than relying solely on trademark law for every voice-cloning dispute.

Voice Is Becoming a Major Digital Asset

AI has transformed voice from something ephemeral into a reusable digital asset.

Previously, hiring a celebrity voice generally required:

• recording equipment;
• the celebrity’s physical participation; and
• repeated studio sessions.

A sufficiently sophisticated AI model can potentially generate unlimited new speech using a relatively small voice sample.

That changes the economics of endorsement.

If businesses could indefinitely clone a celebrity’s voice after obtaining one recording, the commercial market for voice endorsements could be substantially undermined.

Indian courts’ willingness to protect voices therefore has significant commercial implications for actors, singers, voice artists and influencers.

AI Companies and “Consent Once, Use Forever”

Contract drafting will become particularly important in this field.

Suppose an actor records ten sentences for an advertisement.

The agreement permits use of those recordings for one year.

The brand then uses the audio to train an AI model capable of generating the actor’s voice indefinitely.

Can it continue generating new advertisements after the one-year licence expires?

The answer should not be assumed.

The contract would need careful interpretation.

This is why AI-era agreements increasingly need explicit provisions concerning:

• model training;
• synthetic reproduction;
• post-term use;
• deletion; and
• derivative synthetic content.

What If the Celebrity Publicly Posted the Photo or Voice?

Public availability does not automatically mean unrestricted commercial permission.

Posting a photograph publicly on Instagram does not ordinarily mean:

“Anyone in the world may use this photograph to make me endorse their product.”

Similarly, uploading interviews to YouTube does not necessarily amount to consent for third parties to build a synthetic voice clone for commercial exploitation.

Access and permission are different concepts.

What If the AI Output Is Obviously Fictional?

An obviously fictional output may present stronger freedom-of-expression arguments.

For example:

“What if Amitabh Bachchan played Superman?”

accompanied by clearly artistic fan art may raise a different legal analysis from:

“Amitabh Bachchan recommends this crypto platform”

presented as a realistic video.

Relevant questions include:

• commercial purpose;
• likelihood of confusion;
• identifiability;
• context;
• satire;
• artistic expression;
• consumer deception; and
• reputational harm.

Personality Rights and Political Deepfakes

AI-generated depictions of politicians raise additional constitutional considerations.

Political speech receives strong protection, particularly in:

• criticism;
• satire;
• cartoons;
• election commentary; and
• public-interest debate.

At the same time, malicious voice cloning or fabricated videos falsely attributed to a public figure may involve:

• impersonation;
• defamation;
• misinformation;
• election law;
• platform rules; and
• other statutory concerns.

A court must carefully distinguish political criticism from fraudulent impersonation.

Is Every AI Celebrity Impression Illegal?

No.

A blanket rule would be inconsistent with freedom of expression.

Consider four examples.

Example 1: Commercial endorsement

An AI video falsely shows an actor promoting a skincare product.

High legal risk.

Example 2: Voice-cloning service

A paid website lets customers make a famous singer “sing” any song.

Significant personality-rights risk, particularly after Arijit Singh.

Example 3: Newspaper cartoon

A caricature exaggerates an actor’s appearance while commenting on a current controversy.

Stronger speech/parody protection.

Example 4: Academic demonstration

A university researcher demonstrates deepfake technology using limited material for bona fide research.

The legal analysis could involve research, fair dealing, consent, privacy and other factors and would be materially different from commercial exploitation.

Context is essential.

Practical Guidance for AI Companies

AI developers dealing with identifiable real people should consider:

• obtaining explicit consent;
• defining permitted uses;
• restricting celebrity impersonation;
• preventing misleading endorsements;
• clearly labelling synthetic media;
• maintaining provenance metadata;
• providing rapid takedown mechanisms;
• restricting sexually explicit deepfakes;
• preventing fraud and personation;
• respecting contractual restrictions; and
• implementing complaint procedures.

The 2026 IT-rule amendments make synthetic-content governance increasingly important for platforms as well.

Practical Guidance for Brands

A company should not assume that using AI removes the need for an endorsement agreement.

If anything, AI increases the need for contractual permission.

Before generating synthetic celebrity content, businesses should confirm:

• who owns relevant recordings or images;
• whether the celebrity authorised use of their persona;
• whether AI training is contractually permitted;
• whether synthetic reproduction is permitted;
• whether the use implies endorsement;
• whether the resulting content must be labelled; and
• whether trademark, copyright, privacy or consumer law applies.

“AI-generated” is not a defence to unauthorised exploitation.

Practical Guidance for Celebrities

Public figures should increasingly consider protecting their digital identity proactively through:

• trademark registration where appropriate;
• contractual controls over AI use;
• monitoring for deepfakes;
• platform reporting;
• preservation of evidence;
• legal notices;
• intermediary takedown requests; and
• urgent injunction proceedings.

Important evidence includes:

• URLs;
• screenshots;
• screen recordings;
• timestamps;
• advertisements;
• transaction pages;
• subscription pricing;
• social-media handles; and
• evidence showing commercial gain or public confusion.

Frequently Asked Questions

Can AI legally use a celebrity’s face without permission in India?

There is no universal yes-or-no rule for every use.

However, unauthorised commercial exploitation, fake endorsements, deepfakes, merchandising and misleading AI impersonation can violate personality/publicity rights and other laws.

Recent Indian courts have repeatedly restrained such uses.

Can AI clone a celebrity’s voice?

Technically yes.

Legally, unauthorised commercial cloning of an identifiable celebrity’s voice can create serious liability.

The Bombay High Court’s Arijit Singh decision and subsequent cases strongly support protection against such exploitation.

Can I make AI songs using a famous singer’s voice?

Doing so for purely private experimentation is legally different from releasing, monetising or commercially exploiting the song.

Commercial publication of an identifiable cloned singer’s voice without permission presents substantial personality-rights risk.

Can an advertiser use AI to make a celebrity endorse its product?

Doing so without consent presents one of the strongest potential cases for personality-rights infringement, passing off and misleading endorsement.

Are celebrity deepfakes illegal in India?

Some may be unlawful depending upon content and purpose.

Deepfakes involving fraud, impersonation, privacy violations, obscene content or unauthorised commercial exploitation can attract legal consequences.

India also strengthened its IT Rules concerning synthetic media in February 2026.

Can a celebrity stop memes?

Not automatically.

Parody, satire, caricature, commentary and other expressive uses must be balanced against personality rights.

The Karan Johar proceedings expressly recognised this distinction.

Can a celebrity prevent newspapers from using their photograph?

Not simply because they are a celebrity.

Legitimate journalism and public-interest reporting involve important freedom-of-expression principles.

False endorsement is different from news reporting.

Does a disclaimer make an AI deepfake legal?

Not necessarily.

Labelling synthetic content may reduce deception and may be required under regulatory rules, but it does not itself create permission to commercially exploit another person’s identity.

Can ordinary people object to AI cloning their face or voice?

Yes, depending upon the circumstances.

Even where classic celebrity publicity rights are not involved, ordinary individuals may rely on privacy, data-protection, cybercrime and other legal remedies.

The Emerging Indian Legal Position

Several themes can now be identified from Indian jurisprudence.

First, courts increasingly recognise that personality is not limited to a name or photograph.

It may include:

• voice;
• likeness;
• speech style;
• singing mannerisms;
• signature;
• distinctive expressions;
• visual appearance; and
• synthetic recreations.

Second, AI does not create a legal vacuum.

Indian courts have repeatedly applied existing personality-rights doctrine to:

• Generative AI;
• deepfakes;
• voice clones;
• AI chatbots;
• face swapping;
• synthetic media; and
• machine learning.

Third, commercial exploitation remains particularly important.

Where another person profits from a celebrity’s identity without consent, judicial intervention is increasingly likely.

Fourth, personality rights remain subject to constitutional limits.

Parody, criticism, legitimate journalism and artistic expression cannot simply be prohibited because the subject dislikes the content.

Conclusion

Artificial intelligence has changed the practical meaning of celebrity identity.

A person’s face no longer needs to be photographed.

It can be generated.

Their voice no longer needs to be recorded each time.

It can be cloned.

Their endorsement no longer requires their presence.

It can be fabricated.

Indian personality-rights law is consequently undergoing one of its most significant periods of development.

Although India does not yet have a single comprehensive Personality Rights Act, courts have constructed meaningful protection through publicity rights, privacy, passing off, intellectual-property principles and constitutional doctrine.

The direction of recent jurisprudence is unmistakable.

From Anil Kapoor and Arijit Singh to Karan Johar, Sonakshi Sinha, Varun Dhawan, Ravi Kishan, Yuvraj Singh and Khushi Kapoor, Indian courts have increasingly restrained the unauthorised exploitation of recognisable personality attributes through AI and synthetic technologies.

At the same time, the law does not create an absolute monopoly over public identity.

News reporting, parody, satire, criticism and legitimate artistic expression remain important considerations under Article 19(1)(a).

The correct legal principle is therefore not:

“AI can never use a person’s face or voice without consent.”

Nor is it:

“Anything publicly available can be freely used by AI.”

The more accurate position is:

Where AI appropriates an identifiable person’s face, voice, likeness or other personality attributes without authorisation—particularly for commercial exploitation, false endorsement, impersonation, deepfake content or other harmful uses—Indian law increasingly provides strong grounds for judicial intervention.

For the AI industry, this means that technological capability should not be confused with legal permission.

For celebrities, performers and influencers, their digital persona has become an increasingly valuable—and increasingly vulnerable—commercial asset.

And for Indian law, the central challenge will be to protect individuals against synthetic identity exploitation without turning personality rights into an excessive restriction on legitimate speech, creativity, journalism and innovation.

That balance is likely to define the next phase of India’s AI and personality-rights jurisprudence.