Delhi High Court Orders Meta to Remove AI Deepfakes of Kangana Ranaut
According to the Court, Meta Platforms, which owns Facebook and Instagram, must take down all said content that is fake and AI-generated concerning Kangana Ranaut within 72 hours of receiving a copy of the order.
For advertisement on our platform, do call at +91 6377460764 or email us at [email protected].
The order was passed by Justice Vikas Mahajan on October 8, 2026, while hearing the petition filed by Ranaut seeking safeguarding of her personality rights.
The case involves manipulated images, adult deepfake videos, fake speeches, and social media posts purportedly made by the actor without her consent.
The court also directed the defendants to not use her name, image, voice, speech, resemblance or any part thereof in any manner as contrary to the directions of the interim order.
This is significant as it has become easier in recent years to produce convincing videos and photographs of people appearing to do or say things that never occurred, thanks to artificial intelligence.
It can have a negative effect on the reputation of celebrities who are using such content as it can mislead fans and fans can develop their own associations of products and political statements or objectionable material.
But the order is not a nationwide law making it mandatory to delete every deepfake within 72 hours. It is an ad interim order passed in some specific suit.
What Kangana Ranaut Had to Say to the Court:
Ranaut moved the Delhi High Court claiming some websites and social media profiles were putting out content that was falsely using her identity.
As documented in the submission made before the court, the thread in question had the inclusion of a series of AI-generated images, edited videos and fabricated communications in her name.
There were reportedly images of a sexual nature.
Several posts seemed to have Ranaut saying or sharing views that she had not.
According to her attorneys, the review was in violation of her “personality rights, right to identity, dignity and reputation.”
It did not stop there, as it cannot only be the employment of her pictures.
AI can now mimic a person’s voice, face and style of speaking. To a bystander who sees a fake video, it could look real.
These issues are especially concerning when the individual being impersonated is a film actor and an elected representative.
A fake political quote by a celebrity can propagate quickly and has the potential to affect how the celebrity’s views are interpreted.
Consequently, Ranaut had to approach for protection against the further circulation and exploitation of the various other content.
What Exactly Is Meta Required to Do Within 72 Hours?
The court ordered Meta to delete, de-index, and permanently take down the particular infringing content pointed out by Ranaut through the exhibits to the order.
The scope involves content posted on platforms operated by Meta, such as video, reels, posts, accounts and channels.
The clock starts ticking the moment Ranaut delivers a copy of the order to Meta, it said.
This distinction matters.
The timeframe is not just measured from when the judgment was handed down or the date the offensive video was uploaded.
Additionally, the court has ordered the social media giant to submit an affidavit of compliance within two weeks.
A compliance affidavit states in writing the compliance to the directions made before the court.
The order also provides Ranaut a legal recourse to seek action against the material identified instead of relying solely on the standard reporting tools of the social media platforms.
But no-fault order publicly available did not create a lack of conformity Meta deadline.
Government Authorities Also Receive Directions
Although not addressed to Meta exclusively, the directions of the court.
The Ministry of Electronics and Information Technology and the Department of Telecommunications were ordered to act with regard to the websites mentioned in the matter.
The directions were to be issued to the internet service providers and the telecom service providers to block access to those websites.
The directions were given in the same 72 hours period after order was communicated.
Compliance affidavits were also to be filed by the authorities within a fortnight.
The aim was to remove content which could still be accessed on websites on other platforms beyond Meta.
The court also found that the blocking measures could impact websites or accounts that were not mainly responsible for the infringement.
The operators were, therefore, allowed to come to the Court for lifting of the injunction after giving the required undertaking.
This protection applies because directions that are too general could lead to the blocking of lawful material as well as unlawful material.
What Are Personality Rights Under Indian Law?
Those aspects of a person’s personality are protected by personality rights.
Including a person’s name, photograph, voice and those features which serve to identify him.
And for celebrities, these qualities are worth a lot of money.
A popular actor can be ‘paid’ to promote the product, to feature in an ad or to allow the use of his or her image in advertising.
Making it seem like someone else is endorsing the celebrity’s personality could result in monetary or other penalties for the famous individual.
There isn’t a single law in India that covers the entire personal areas of individual rights.
Rather, protections have been added to constitutional principles, privacy law, intellectual property principles and civil remedies.
The right to privacy and dignity can be drawn in case of personal identity.
Commercial misuse may also give rise to passing off, false endorsement and other claims depending on the facts.
In the case of Ranaut, the High Court established a prima facie foundation for the protection of her personality traits from the alleged use.
The findings of the court were made during the interim time, prior to the conclusion of the lawsuit.
What Makes Deepfakes Unique from Other Photos You Can Edit
We’ve been manipulating video and photos for decades.
Artificial Intelligence has made realistic manipulation easier and faster.
A deepfake can create a person saying things they never said, being in a situation they were never in, or even creating sexually explicit images of them.
Others are edited for fun or parody.
Some others are made to trick the audiences, slander someone’s reputation, pretend to be someone else or commit scams.
The more significant of the two problems is the fact that the manipulative material does not necessarily look so obviously false.
A false endorsement is when consumers might believe that a celebrity endorses a product.
Bim your reputation or (sometimes deliberately) make a person say or blog something false.
Sexual deepfakes intended for non-consensual sharing can be especially harmful because they appropriate the victim’s identity for sexual content.
It must also take account of what the image depicts, its intended purpose and possible impact.
What was the Court’s statement on fictitious speeches and vulgar material?
Justice Vikas Mahajan assessed the material furnished by Ranaut and felt that the alleged injury to her image and personality was grave enough to call for interim relief.
The court mentioned that the material was of an obscene, sexual, or pornographic nature and was supposed to show Ranaut through AI or others.
It also involved false statements and changed recordings.
The court found that fake statements in live audio and video stream can cause an average viewer to believe Ranaut that had actually made the statement.
This gets more serious when the fake material uses the individual’s public position to enhance the statement’s credibility.
So, this order is not dealing with only unauthorised commercial endorsements.
