Delhi HC Orders 72-Hour AI-Deepfake Takedowns | TLY

AI Regulation Tracker  /  Enforcement

Delhi High Court Orders 72-Hour AI-Deepfake Takedowns and 3-Day Registrar Removals to Protect Ravi Kishan's Persona

Regulatory summary: Ravindra Shukla alias Ravi Kishan v. Ashok Kumar (John Doe) and Others, CS(COMM) 680/2026, reported as 2026 LiveLaw (Del) 632, is an interim order of the Delhi High Court passed by Justice Jyoti Singh on July 2, 2026.

In Ravindra Shukla (Ravi Kishan) v. Ashok Kumar, the Delhi High Court granted an ex parte injunction against AI, generative AI, and deepfake misuse of the actor and Member of Parliament's identity, and imposed short-clock takedown duties: domain registrars must remove listed URLs within three days, and Meta, Google, and X must act within 72 hours of notice. The interim order took effect on July 2, 2026.

Primary source

Delhi High Court Orders 72-Hour AI-Deepfake Takedowns and 3-Day Registrar Removals to Protect Ravi Kishan's Persona regulation briefing
The Leveraged Years AI Regulation Tracker

Key takeaways

  • The Delhi High Court extended personality-rights protection to AI and deepfake misuse and attached enforceable short-clock takedown duties to it. Rather than a general restraint alone, the order sets concrete deadlines: three days for registrars and website operators to remove listed URLs, and 72 hours for Meta, Google, and X to act on identified content after notice. That converts a rights declaration into an operational compliance timetable.
  • Domain name registrars and hosting providers; large intermediary platforms named in the order and others like them; advertisers and agencies using AI-generated or cloned likenesses of public figures; celebrities, athletes, and other public figures and their counsel; and legal and trust-and-safety teams that manage takedown requests in India.
  • Status: Interim order passed and in effect.
  • If you are a platform or registrar, map your notice-to-removal process against the 72-hour and three-day windows and fix any step that cannot meet them. If you are an advertiser or agency, add a likeness-clearance gate that blocks any AI-generated or cloned depiction of a real person until written rights are on file.
DateJurisdictionRuleAffected professionalsStatus or effective date
2026-07-09IndiaThe Delhi High Court extended personality-rights protection to AI and deepfake misuse and attached enforceable short-clock takedown duties to it. Rather than a general restraint alone, the order sets concrete deadlines: three days for registrars and website operators to remove listed URLs, and 72 hours for Meta, Google, and X to act on identified content after notice. That converts a rights declaration into an operational compliance timetable.Domain name registrars and hosting providers; large intermediary platforms named in the order and others like them; advertisers and agencies using AI-generated or cloned likenesses of public figures; celebrities, athletes, and other public figures and their counsel; and legal and trust-and-safety teams that manage takedown requests in India.Interim order passed and in effect. Pronounced July 2, 2026, reported as 2026 LiveLaw (Del) 632, and covered in legal and trade press from July 6 to July 9, 2026. The plaintiff was directed to comply with Order 39 Rule 3 of the Code of Civil Procedure within the time set by the court. The matter remains pending for further hearing.

Frequently Asked Questions

Is this a final ruling against the defendants?

No. It is an ex parte ad interim injunction, meaning a provisional order passed without first hearing the defendants. It is enforceable now but the case remains pending for further hearing, and the plaintiff was directed to serve the order under Order 39 Rule 3 of the Code of Civil Procedure.

What exactly are the takedown deadlines?

Website operators and domain name registrars must remove the listed infringing URLs within three days. Intermediary platforms, including Meta, Google, and X, must take down identified content within 72 hours of receiving notice from the plaintiff.

Does the order specifically cover AI and deepfakes?

Yes. The court restrained use of the actor's name, image, likeness, and voice through any medium, expressly including artificial intelligence, generative AI, machine learning, deepfakes, social media, and virtual platforms.

Are advertisers who use AI likenesses at risk?

Yes. The order treats unauthorised commercial exploitation of a person's persona as actionable and names AI-generated likenesses. Using a synthetic depiction or cloned voice of a real person without written rights is the kind of use the court restrained.

Does this bind platforms that did not create the content?

Yes, for removal. Registrars and named intermediaries are not accused of creating the content, but they carry direct duties to remove or take down identified material within the court's deadlines once notice is served.

Browse the full AI Regulation News tracker

Informational analysis for working professionals, not legal advice. Confirm how any rule applies to your situation with qualified counsel.