Deepfakes & Personality Rights: Examining the Role of Copyright, Trademark and Passing Off in India
Artificial Intelligence has made it possible to create photographs, videos and audio recordings that look and sound almost exactly like real people. One of the most concerning examples of this technology is the deepfake. A deepfake can make it appear as if a person has said something they never said, attended an event they never attended, or promoted a product they have never even heard of.
Introduction
Artificial Intelligence has made it possible to create photographs, videos and audio recordings that look and sound almost exactly like real people. One of the most concerning examples of this technology is the deepfake. A deepfake can make it appear as if a person has said something they never said, attended an event they never attended, or promoted a product they have never even heard of.
The problem becomes more serious when the identity of a person is used without their permission. A celebrity's face or voice can be used to advertise a product, while an ordinary person's photograph can be manipulated and circulated on social media. For example, in the Arijit Singh v Codible Ventures LLP case, AI was allegedly used to recreate Arijit Singh's voice, while his likeness was also used in advertising in a way that suggested his endorsement without authorisation.1 Similar misuse can also affect ordinary individuals, whose photographs may be manipulated and circulated online without their consent.
In both situations, the person may suffer loss of reputation, privacy and dignity.
In the case of celebrities, there can also be a direct financial loss because their name, image and voice have commercial value. Their identity is often connected with advertisements, endorsements, films and other commercial activities. The Delhi High Court recognised this concern in Anil Kapoor v Simply Life India, where it protected the actor against unauthorised commercial use of his name, image, voice and other aspects of his persona, including misuse through AI and deepfake technology. The Court also recognised that such unauthorised use could cause financial harm to the celebrity.2
This raises an important legal question: If artificial intelligence creates a fake version of a person's identity, what legal remedy does that person have in India?
India does not currently have one separate law dealing specifically with personality rights or AI-generated deepfakes. Instead, protection has developed through different areas of law,including copyright, trademark law, passing off, privacy and constitutional rights. Courts have also played an important role in recognising personality and publicity rights.
The World Intellectual Property Organisation (WIPO) has also highlighted this problem. In its 2026 article on AI deepfakes and Indian cricketers, WIPO discussed how generative AI can be used to create highly realistic images of individuals without their consent and how such use can interfere with their personality and publicity rights.3 In its discussion of AI voice cloning, WIPO observed that Indian courts are increasingly using existing legal principles relating to personality rights and intellectual property to deal with AI-generated misuse. The decision involving singer Arijit Singh is particularly important because it directly dealt with the unauthorised use of AI to reproduce a celebrity's voice and personality.4
What exactly is a deepfake?
In simple terms, a deepfake is digitally manipulated content created using artificial intelligence in which a person's face, voice or other characteristics are placed into content that they did not actually create or participate in. For example, imagine a famous actor appearing in a video where he seems to endorse a particular betting application. The actor never made the video, but AI has been used to reproduce his face and voice. To an ordinary viewer, the video may appear genuine.
Similarly, a singer's voice can be copied by an AI system and made to sing words or songs that the singer never recorded. This is where deepfakes become different from ordinary photographs or videos. The technology does not merely copy an existing image; it can create new content using a person's identity. This creates a difficult legal situation because traditional intellectual property laws generally protect things such as creative works, trademarks and commercial goodwill. They do not specifically provide a single right over a person's entire identity.
Personality Rights and Right of Publicity
Personality rights essentially protect the identity and persona of an individual. They may include a person's name, photograph, face, voice, signature, likeness, distinctive mannerisms and other characteristics by which the person can be identified.5 The commercial side of these rights is commonly referred to as the right of publicity. In simple words, it means that a person should have some control over the commercial use of their identity. Indian courts have gradually recognised these rights even though India does not have a separate Personality Rights Act.
An important early case was D.M. Entertainment Pvt Ltd v Baby Gift House. The Delhi High Court recognised that a celebrity has an interest in controlling the commercial exploitation of his identity and that unauthorised commercial use could give rise to claims involving publicity rights, passing off and false endorsement.6
The importance of this principle becomes much clearer in the age of AI. Earlier, someone might have printed a celebrity's photograph on a product without permission. Today, the same person can use AI to create a completely new advertisement featuring the celebrity's face and voice. For example, an AI tool could create a video showing a famous actor promoting a skincare product, using the actor’s face and an AI-generated version of their voice, even though the actor never actually promoted or endorsed the product. This can make viewers believe that the celebrity genuinely supports the product.
The basic problem, however, remains the same: someone else is commercially benefiting from a person's identity without permission.
Copyright: Can it protect against deepfakes?
Copyright is one of the first areas of intellectual property law that comes to mind when dealing with AI-generated content. However, copyright does not directly protect a person's identity.
Under Section 14 of the Copyright Act 1957, copyright gives certain exclusive rights to the owner of a protected work. These rights cover works such as photographs, films, sound recordings and other artistic and literary works.7
Therefore, copyright can become relevant where a deepfake uses an existing copyrighted work. For example, suppose a photographer owns the copyright in a photograph of a celebrity. Someone takes that photograph, uses AI to manipulate it and then commercially distributes the resulting content. The copyright owner may have a claim in relation to the unauthorised use of the copyrighted photograph.
Similarly, if an existing song or sound recording is taken and manipulated as part of an AI-generated video, copyright issues may arise.
However, there is a major limitation.
Suppose an AI system creates a completely new video of a celebrity using only a few photographs as references. The final video may not directly reproduce any particular copyrighted photograph. The real problem is that the person's face and identity have been recreated without permission. This shows one limitation of relying only on copyright, as copyright protects creative works, while the misuse of a person's identity is more directly connected with personality and publicity rights. This issue was seen in Yuvraj Singh v. Ashok Kumar & Ors. (2026),8 where The Delhi High Court held that Yuvraj Singh has the right to protect his name, image, voice, likeness and other personality attributes from unauthorised use. The Court found that AI-generated posts and deepfake videos falsely depicting him could damage his reputation and goodwill. It also held that unauthorised commercial use of his personality attributes could amount to passing off and reduce their commercial value. Accordingly, the Court granted an interim injunction restraining the defendants from using his personality through AI, generative AI, deepfakes, face-swapping and similar technologies without his consent.
In such a situation, copyright alone may not provide a complete solution because copyright protects the creative work and not the person's identity as such.
This distinction is particularly important in the case of AI-generated deepfakes.
Protection of performers
The Copyright Act may nevertheless provide additional protection to performers. Section 38B recognises certain moral rights of performers and protects them against distortion, mutilation or other modification of their performance which is prejudicial to their reputation.9
This becomes relevant when AI is used to manipulate an existing performance. Therefore, copyright law can form one part of the legal protection, but it cannot by itself become a complete law against deepfakes.
Trademark Law and the Use of a Person's Name
Trademark law provides another possible remedy, particularly when a person's name has commercial value.
Section 2(1)(m) of the Trade Marks Act 1999 includes a name within the definition of a "mark".10 A celebrity may also register their name or other distinctive identifiers as trademarks where the requirements of trademark law are satisfied.
For example, if a celebrity has a registered trademark in their name and another person uses that name in connection with goods or services in a manner that amounts to infringement, the Trade Marks Act may provide a remedy.
The decision in Jaikishan Kakubhai Saraf alias Jackie Shroff v The Peppy Store is useful in understanding this area.11 The Delhi High Court held that a celebrity has protectable personality and publicity rights over attributes such as their name, image, voice, likeness and other distinctive characteristics. The Court observed that unauthorised commercial use of these attributes can amount to infringement of personality rights and can result in unfair commercial gain to others. The Court therefore restrained several defendants from using Jackie Shroff’s name, voice, image and other personality attributes for commercial purposes without his consent. It also restrained the commercial use of an unlicensed AI chatbot that imitated his persona. However, the Court recognised that personality rights must be balanced with legitimate artistic expression and freedom of expression.
However, trademark law also has its limitations. A person's identity can be misused even when no trademark is involved. For example, a deepfake video may show a celebrity making a political statement or saying something offensive. There may be no registered trademark being infringed at all. Therefore, trademark law can help where the identity is being commercially used as a source or indication of goods or services, but it cannot cover every kind of deepfake.
Passing Off and False Endorsement
Passing off is particularly useful in cases involving celebrities because it can provide protection even when a person does not have a registered trademark. The basic idea behind passing off is quite simple: one person should not create a false impression that their goods, services or activities are connected with someone else's reputation or goodwill.
Consider a situation where an AI-generated video shows a famous actor saying:
«"I personally recommend this product."»
If the actor has never made such a statement, the video may give consumers the false impression that the actor has endorsed the product. This is where the idea of false endorsement becomes important. In D.M. Entertainment, the Delhi High Court recognised the importance of protecting a celebrity against unauthorised commercial exploitation of his identity.12 The same principle has subsequently been used in several personality-rights cases.
Passing off therefore becomes particularly useful where a deepfake is being used to make money by creating a false association with a celebrity. However, even passing off has limitations. A deepfake does not always involve commercial activity. Someone may create a fake video simply to insult, defame or harass an individual. In such a situation, passing off may not be the most appropriate remedy. This shows why deepfakes cannot be dealt with through one branch of intellectual property law alone.
The Arijit Singh Case: A Major Development
One of the most important recent developments is Arijit Singh v Codible Ventures LLP. The case is particularly relevant because it directly involved the use of artificial intelligence to reproduce a person's voice and likeness.
Arijit Singh alleged that his voice was being recreated through AI tools without his permission. His name and likeness were also being used in a manner that suggested an association with certain activities and events.The Bombay High Court recognised the importance of protecting Singh's personality rights. The protection was not limited to his photograph or name. It extended to characteristics including his voice, vocal style, manner of singing and other distinctive aspects of his persona.13
WIPO's analysis of the case describes it as an important Indian decision dealing with the intersection of generative AI, intellectual property and personality rights. WIPO also noted the Court's concern that AI tools capable of recreating a celebrity's voice could be misused in ways that harm both the person's economic interests and reputation.14 The case is particularly significant because a voice is not necessarily a traditional copyright work. Yet, for a singer such as Arijit Singh, his voice is one of the most recognisable parts of his professional identity.
The decision therefore shows that the law may have to look beyond traditional categories of intellectual property when dealing with AI.
Personality Rights, Privacy and Article 21
Deepfakes also raise questions that go beyond commercial loss. The Supreme Court's decision in Justice KS Puttaswamy (Retd) v Union of India recognised privacy as a fundamental right under Article 21 of the Constitution.15 The right to privacy includes concerns relating to dignity, autonomy and control over personal aspects of one's life. This becomes particularly important when the deepfake is not being used to sell a product but to humiliate or harm an individual.
For example, an AI-generated video may show a person making offensive statements or place their face onto sexually explicit content. Even if no commercial benefit is involved, the individual can suffer serious emotional and reputational harm. Recent Indian decisions show that courts are increasingly recognising this broader dimension. WIPO's 2026 discussion of AI deepfakes notes the Delhi High Court's approach of connecting protection against AI-generated misuse with the individual's right to live with dignity under Article 21. 16 This is important because it shows that the deepfake problem is not merely an IPR problem. It can also be a question of privacy, dignity and personal autonomy.
What about ordinary individuals?
Most discussions on personality rights focus on celebrities because their identities have obvious commercial value. However, deepfakes can affect ordinary individuals just as seriously. A normal person's face can be taken from social media and used to create a fake video. Their voice can be cloned and used in a fraudulent phone call. Their photograph can be manipulated and circulated online.
The difference is that a celebrity may have an established legal and commercial identity, while an ordinary person may find it more difficult to establish a claim under passing off or trademark law.
This raises an important question: Should protection against identity misuse depend upon whether the person is famous? The answer should ideally be no.
WIPO's discussion of the Arijit Singh case also recognises this broader concern, noting that legal protection should not be limited only to celebrities because every individual has an interest in protecting their personality and privacy.
Is Existing Indian Law Enough?
Existing Indian law can certainly provide remedies against some forms of deepfake misuse. Copyright may apply when copyrighted material is copied. Trademark law may help where a protected name or mark is commercially misused. Passing off may help where false endorsement or an association with someone's goodwill is created. Personality rights can provide protection against unauthorised commercial exploitation of identity. Constitutional privacy and dignity principles may also become relevant.
However, the present framework still has gaps. The biggest problem is that these protections are spread across different laws and judicial principles. There is no single law that clearly answers questions such as:
Who owns the rights in an AI-generated replica of a person's face or voice?
Does a person have a right over their voice even if it has not been registered as a trademark?
What happens when a deepfake is created for non-commercial purposes?
What should happen when the person is an ordinary individual rather than a celebrity?
Who should be responsible when a platform hosts thousands of AI-generated deepfakes
How quickly should such content be removed?
What happens when the person creating the deepfake is anonymous?
These questions become even more difficult because deepfake technology is developing much faster than the law.
Need for Stronger Legal Protection
India does not necessarily need to create a completely separate law for every new AI technology. However, there is a strong need for clearer rules dealing with unauthorised digital replicas and identity misuse. Some possible solutions are:
1. Clear recognition of personality rights:A future legal framework should expressly recognise a person's right to control the commercial use of their name, face, voice, likeness and other distinctive identity features. This would provide clearer protection against their unauthorised use through AI.
2. Faster removal of harmful deepfakes: A simple and time-bound procedure should be established for reporting and removing harmful deepfake content from online platforms. Platforms should be required to respond quickly when such content violates a person's rights.
3. Effective remedies against misuse: The law should provide effective remedies such as injunctions, compensation and penalties against those who create or commercially exploit unauthorised deepfakes. This would make the protection more effective in practice.
4.Balance with freedom of expression:The law must also recognise that not every use of someone's image is unlawful. Satire, criticism, parody, journalism, artistic expression and legitimate public-interest content should be protected so that stronger personality rights do not unnecessarily restrict freedom of expression. The aim should therefore not be to prevent people from using AI to create content. The aim should be to prevent deceptive, harmful or unauthorised exploitation of another person's identity.
Conclusion
Deepfakes have created a situation where a person's identity itself can be copied and manipulated. This creates a challenge for traditional intellectual property law because copyright, trademark and passing off were not originally designed for a world in which artificial intelligence could reproduce a person's face or voice within seconds.
Indian courts have nevertheless shown that existing legal principles can be adapted to new technology. From D.M. Entertainment to Anil Kapoor, Jackie Shroff and, more recently, Arijit Singh, the courts have increasingly recognised that a person's identity can have both personal and commercial value.17
The developments discussed by WIPO also show that AI is not making personality rights irrelevant; rather, it is making them more important.18 At present, India's approach is therefore best described as a combination of different legal protections rather than one comprehensive solution. Copyright can protect the underlying creative work, trademark law can protect commercial identifiers, passing off can address false association and endorsement, while personality and privacy rights can protect the individual more directly. Ultimately, the challenge is to ensure that technological progress does not come at the cost of an individual's identity, dignity and control over their own persona. India may therefore need clearer statutory protection in the future, particularly as AI-generated deepfakes become easier and cheaper for ordinary users to create.
[1] Dipak G. Parmar, 'AI voice cloning: how a Bollywood veteran set a legal precedent' WIPO Magazine (17 April 2025).
[2] Anil Kapoor v Simply Life India & Ors., 2023 SCC OnLine Del 6914.
[3] Tania Sebastian and Saraswathy Vaidyanathan, ‘Combatting AI deepfakes to safeguard Indian cricketers’ image rights’, WIPO Magazine (14 July 2026).
[4] Dipak G Parmar, 'AI voice cloning: how a Bollywood veteran set a legal precedent' WIPO Magazine (17 April 2025).
[5] World Intellectual Property Organization (WIPO), Personality Rights, WIPO IP and Innovation Researchers Forum, available at WIPO.
[6] D.M. Entertainment Pvt Ltd v Baby Gift House 2010 SCC OnLine Del 4790.
[7] Copyright Act 1957, s 14
[8] Yuvraj Singh v. Ashok Kumar & Ors., Delhi High Court, 29 July 2026.
[9] Copyright Act 1957, s 38B.
[10] Trade Marks Act 1999, s 2(1)(m).
[11] Jaikishan Kakubhai Saraf alias Jackie Shroff v The Peppy Store 2024 SCC OnLine Del 3664.
[12] D.M. Entertainment Pvt Ltd v Baby Gift House 2010 SCC OnLine Del 4790
[13] Arijit Singh v Codible Ventures LLP (Bombay High Court, 2024).
[14] WIPO's discussion specifically identifies the protection of Singh's name, voice, likeness, vocal style, manner of singing and other personality attributes in the context of AI-generated content.
[15] Justice KS Puttaswamy (Retd) v Union of India (2017) 10 SCC 1.
[16] WIPO, 'Combatting AI deepfakes to safeguard Indian cricketers' image rights' (2026).
[17] D.M. Entertainment (n 2); Anil Kapoor v Simply Life India 2023 SCC OnLine Del 6914; Jaikishan Kakubhai Saraf (n 6).
[18] WIPO, 'AI voice cloning: how a Bollywood veteran set a legal precedent' (n 1); WIPO, 'Combatting AI deepfakes to safeguard Indian cricketers' image rights' (n 11)