Deepfaked? How Indian Courts Protect Your Face, Voice and Name, and What a Brand Should Do First

The video looks like your founder. It sounds like her. It tells viewers to put money into a trading scheme, and it has been shared forty thousand times since Tuesday. Your customers are messaging to ask whether it's real.

India has no law called a personality rights statute. Yet in 2026 alone, the Delhi and Bombay High Courts have issued a string of orders against AI deepfakes, voice clones and fake endorsements, and the government has tightened platform takedown times to hours. This guide explains how personality rights in India are being protected against deepfakes, where the courts are now drawing the line, and what a business should do in the first few days.

Quick answer

There is no statute on personality rights in India. Courts protect a person's name, image, voice and likeness through passing off, the right to privacy and dignity under Article 21, and, where they apply, the Trade Marks Act and the moral rights provisions of the Copyright Act. Courts routinely order platforms to remove deepfakes and disclose who posted them. Since August 2026, though, the Delhi High Court has refused blanket bans, asked for evidence of commercial misuse, and protected satire. Separately, IT Rules amended in February 2026 require platforms to act on court orders within three hours.

What "personality rights" means when there's no statute

A personality right is the right to control commercial use of who you are: your name, face, voice, signature style, catchphrases. India has never legislated for it. The courts built it from pieces.

• Passing off stops someone from making the public believe you endorse something you don't.
• Article 21, as read in the Supreme Court's privacy cases, protects dignity and control over one's own identity.
• The Trade Marks Act applies where the name, logo or catchphrase is registered.
• The Copyright Act gives performers moral rights over distortion of their performances, which the Bombay High Court relied on in its 2024 order protecting singer Arijit Singh against AI voice cloning.

The Delhi High Court's 2023 order for Anil Kapoor was the first to name AI, deepfakes and face-morphing expressly. Since then the pattern has spread fast.

The 2026 orders, including Ravi Kishan

On 2 July 2026, Justice Jyoti Singh granted actor and MP Ravi Kishan an ex parte interim injunction in Ravindra Shukla alias Ravi Kishan v. Ashok Kumar & Ors. It restrains use of his name, image, voice, likeness and persona through AI, generative AI, machine learning and deepfakes, in any format. The material before the court included deepfake videos, cloned-voice advertisements and statements he never made. Platforms including Meta, Google and X were directed to take down notified URLs, reportedly within 72 hours.

It was one of many. A selection from this year:

The orders share a toolkit: an injunction against unknown defendants ("Ashok Kumar" or John Doe), URL-by-URL takedown directions to platforms, disclosure of the account holders' details, and liberty to notify new URLs as they appear.

The pushback: what courts have started refusing

This is the part most coverage misses.

On 1 July 2026, Justice Subramonium Prasad refused MP Raghav Chadha a blanket ban on deepfakes and satirical content, observing that "public figures must accept being at the receiving end of satirical humour". Only six vulgar posts were ordered down.

In August and September, Justice Anup Jairam Bhambhani went further. Hearing the Aaradhya Bachchan matter, he remarked that personality rights were "acquiring amoebic proportions" and asked whether defamation was being dressed up as intellectual property. In the Janhvi Kapoor and Vivek Oberoi suits on 3 September, he refused a blanket takedown of 6,884 URLs, ordered sexual content removed, and appointed amici curiae to help the court sort the rest. On 18 September, in actor Meenakshi Chaudhary's suit, he asked: "What is the evidence that they made money on your personality?"

Earlier, in May, the court held that a single academic achievement, however impressive, does not by itself create a personality right, while still restraining defamatory content about the student concerned.

The direction is clear. Sexual and obscene deepfakes, impersonation and fraud get fast relief. Broad claims that sweep in satire, fan content and criticism now face scrutiny, and commercial claims need evidence of commercial use.

Founders and companies: why the trade mark matters

You don't need to be a film star. Founders and companies have obtained relief where they could show a reputation and a commercial harm:

• Global Health (Medanta) and Dr Naresh Trehan (January 2025): deepfake medical-advice videos using the hospital's mark and its chairman's face.
• Ankur Warikoo and his company (May 2025): deepfake investment-advice scams.
• Aman Gupta (May 2026) relied on his registered trade marks.
• India TV (August 2026) obtained a final decree covering its registered marks alongside Rajat Sharma's persona.

Notice the pattern. Where the company's registered mark appears in the fake, the brand can sue in its own right for infringement and passing off, and the founder's personality claim rides alongside. That is a much firmer footing than personality rights alone, particularly now that courts are asking harder questions about them.

The new IT Rules on synthetic content

The IT (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2026, notified on 10 February and in force from 20 February 2026, deal with deepfakes directly.

• They define "synthetically generated information": audio, visual or audio-visual content artificially created or altered so that it appears real. Text alone, routine editing and accessibility tools are excluded.
• Platforms must label synthetic content prominently and, where technically feasible, embed provenance metadata. Large platforms must ask users to declare synthetic uploads and verify those declarations.
• Takedown on a court order or government notice must now happen within 3 hours, down from 36.
• User complaints about intimate images or impersonation must be acted on within 2 hours.

These rules don't create a personality right, and they don't give you damages. They do make the platform route much faster, and a platform that ignores them risks its legal protection for user content.

Nothing else is on the statute book yet. In August 2026 the government told Parliament it relies on existing IT Act and criminal provisions, and a parliamentary committee preferred amending the IT Act and Rules to a standalone deepfake law.

What courts want to see

Drawn from the orders above, a strong application usually has:

1. A URL-by-URL schedule, sorted into categories: sexual or obscene; commercial (ads, merchandise, fake bookings, apps); impersonation or fraud; and anything that might be satire or comment.
2. Captures of each item, with the platform, handle, seller or domain registrar identified.
3. Proof of reputation: career record, public profile, media coverage, and registered trade marks.
4. Proof of commercial misuse: ads, payment links, product listings, subscription funnels. View counts alone are no longer enough.
5. Evidence of confusion or harm: customer messages, complaints, anyone who paid money.
6. Urgency, to justify relief without first notifying the defendant.

And, increasingly important, a narrow prayer. Leave out plainly satirical and news content. Asking for everything now risks getting less.

The first 48 hours: a response plan

A practical sequence, not a substitute for advice on your facts.

Hour 0–6: preserve. Capture every instance with URL, date, time, handle and view count. Save the video files. Screenshot any payment page, phone number or link. Do this before reporting, because reported content disappears, and your evidence goes with it.

Hour 0–12: warn your audience. Post a short, factual notice on your own channels that the video is fake and that you never ask for money this way. NSE did exactly this in 2024 when deepfakes of its chief executive circulated.

Hour 2–24: report to platforms. Use each platform's impersonation reporting route and cite the IT Rules. Where there is fraud, file with the national cybercrime portal (cybercrime.gov.in) or call 1930, and keep the acknowledgement.

Day 1–2: check your marks. Is your brand name registered? Your logo? The founder's name, if it is used commercially? If not, file now. A registration won't take effect overnight, but a pending application is better than nothing, and it builds the footing for the next incident.

Day 2 onwards: decide on court. If the material is spreading, making money or causing real harm, a suit seeking a dynamic injunction against John Doe defendants is the tool the courts have been using. Go in with the evidence file above and a prayer you can defend.

How MYCrave can help

MYCrave Consultancy & Services works on the brand side of this problem, which is where most businesses are exposed:

• Trademark search and filing for your brand name, logo, and where appropriate a founder's name or signature phrase, in the classes that matter;
• Watchdog monitoring for misuse of your marks online and in the trade marks journal;
• An evidence file built to the standard courts are now asking for: URL schedules, categories, captures and proof of your marks;
• Platform complaints framed around your registered rights, and coordination with litigation counsel when a suit is needed.

Registration is the part you can do before it happens

You can't stop someone from making a deepfake. You can decide in advance how quickly you respond and how strong your position is when you do. The businesses with the cleanest outcomes this year had registered marks, a public record of reputation, and evidence gathered before it vanished.

Frequently asked questions

Keep reading