To Stop AI Cloning, India's Courts Build Personality Rights—Case by Case

Recently, actor Shruti Haasan became the latest public personality to seek court protection from unauthorised use of her identity online. She pointed out that her images and persona were being used to generate AI deepfake videos, manipulated photographs, fake endorsements, and sexually suggestive content.

Websites sold merchandise in her name, chatbots let users interact with AI versions of her, and other platforms carried AI voice simulations and fabricated endorsements she never authorised.

Haasan approached the Bombay High Court seeking protection of her personality and publicity rights. The court found a "very strong prima facie case" and said the continued circulation of deepfake material could cause irreparable injury because it can be replicated and circulated indefinitely.

Earlier, celebrities including Anil Kapoor, Aishwarya Rai Bachchan, Hrithik Roshan, Karan Johar, and Sonakshi Sinha secured interim injunctions from high courts. Artists like Arijit Singh and Asha Bhosle have also filed cases against AI-cloned voices.

The cases reflect a growing problem that is becoming harder to address through individual court orders. AI can reproduce a face, voice, and mannerisms at scale, and India has no statute that comprehensively defines personality rights or sets out a dedicated framework for AI-generated identity misuse.

"Instead, that person will have to rely on different legal provisions, such as privacy laws, laws regarding impersonation, identity theft, or fraud and personality rights," Mariya Sajan, a corporate lawyer from Kerala, tells AIM. "If somebody needs those rights to be enforceable, they will have to approach a court and establish how the rights are being violated."

A Patchwork of Protections

Indian personality rights cases hinge on constitutional protections, common law, and IP law, with courts relying on Article 21, which protects life, liberty and dignity.

The Haasan case itself invoked Article 19(1)(a), Article 21, the Copyright Act, and common law protections against passing off and unjust enrichment.

Senior Advocate Pavan Duggal, a Supreme Court of India advocate and cyberlaw expert, tells how courts have adapted these protections to deepfakes and digital replicas.

"What we are witnessing is not the emergence of a new right, but the improvisation of an old one under new pressure," he notes, adding that the Delhi High Court and Bombay High Court have done "genuinely creative work" stitching Article 21's dignity and privacy guarantees with the common law tort of passing off.

But those protections remain case-specific. "There is no codified definition of what a personality or publicity right in India actually covers, no statutory threshold for what counts as an unauthorised 'AI-generated' likeness, and no settled position on whether these rights survive a person's death," Duggal observes.

Two Systems

For a celebrity with strong legal representation, a High Court injunction can offer rapid intervention. For an ordinary person, the process is more elongated.

Duggal said a victim should first preserve evidence like screenshots, URLs, handles, dates, and times, since once content disappears online, the evidence does too.

They do have some tools at their disposal now. Under the IT Rules amended in February 2026, a platform must comply with a court or government order for takedown requests within three hours, and act on non-consensual intimate imagery, including deepfakes, within two hours of a complaint, subject to the applicable rule and procedure.

Victims can complain to the platform's Grievance Officer, then the Grievance Appellate Committee if the complaint is not resolved, and to the National Cyber Crime Reporting Portal or the 1930 helpline. Police can also register cases under IT Act Sections 66C, 66D, 66E, 67, and 67A.

So far, it has mostly been celebrities and other public figures, including Union Minister Nitin Gadkari, who have .

A Tech Law Forum, NALSAR study published in March 2026 identified 25 Indian deepfake-takedown cases from 2023 through February 9, 2026. Its dataset found only two non-celebrity petitioners to obtain after sexually explicit deepfakes were circulated.

Duggal calls the judicial route "effective, but slow, resource-intensive," often inaccessible without money or legal support. John Doe orders—against unknown parties—can help against unidentified infringers, but the victim still has to initiate the legal process.

A Harder Liability Question

The 2026 IT Rules place additional obligations on intermediaries dealing with synthetically generated information. The government says intermediaries must ensure clear labelling and traceable metadata for permissible AI-generated content and deploy reasonable technical measures to prevent unlawful synthetic content.

Those rules do not settle what happens when a generative AI system itself produces a harmful representation after a user's prompt.

Duggal's proposed solution is "non-delegable obligations" for AI companies whose systems enable identity replication.

Under that approach, the company which itself controls the model and commercially benefits from its operation is liable. However, whether an AI company is an intermediary or a participant remains unsettled.

What Happens When the Person is Deceased?

Can a person's voice, face ,or persona be digitally recreated after death, and can their family stop it?

Indian courts have taken a restrictive approach. In Deepa Jayakumar v. A.L. Vijay(2020), the Madras High Court, ruling on a proposed biopic of former Tamil Nadu chief minister J Jayalalithaa, held that her personality rights, along with her privacy and reputation, were extinguished on death and did not pass to her heirs.

The Delhi High Court addressed the issue again in 2023. In Krishna Kishore Singh v. Sarla A. Saraogi, over a film on actor Sushant Singh Rajput's life, it held that privacy, publicity and personality rights are not heritable, though defamation of a deceased person's name may still be actionable.

Protected Speech 

Any new law would also have to operate alongside Article 19(1)(a), which protects free speech.

Duggal says legislation should distinguish commercial exploitation and deceptive impersonation from protected expression like criticism, satire, parody, news, and academic work.

Courts have already made room for that. In Digital Collectables v. Galactus Funware (2023), the Delhi High Court recognised that uses involving lampooning, satire, parody, art, scholarship, music, academics, and news can receive protection under Article 19(1)(a).

That distinction becomes particularly important when AI content involves a public figure, because a false endorsement can use the same identity markers as a parody or other commentary.

Duggal believes a useful test would include whether a reasonable person is likely to believe the content is genuine and whether the person's identity is being commercially exploited without consent. He also argues for proportionate remedies, including targeted takedowns rather than blanket restrictions and mechanisms allowing creators to challenge wrongful removals.

Rules Without a Personality Rights Law

Government policy has focused on platform obligations and disclosure, rather than creating a standalone personality rights regime.

The result is a legal framework that continues to develop through individual disputes. What remains unresolved is the scope of what personality rights cover, how ordinary victims can access remedies, when AI companies can be held responsible, whether rights survive death, and how far personality protection can extend without restricting protected speech.

Duggal asserts a statute would need to define personality and publicity rights, fix obligations for AI companies, deter non-consensual impersonation, and preserve legitimate speech. "Until that legislative architecture exists, we will keep seeing excellent judicial carpentry on a house with no foundation."

Key events like the swift institutional reversal under the NALSAR correction, the qualification around takedown timelines, and the distinction between Duggal's proposed “non-delegable obligations” and existing Indian law offer a pathway for courts to discuss the the scope of the interim protections till the Parliament is stirred into action.