Delhi High Court Orders Takedown of Six Vulgar AI Posts Targeting Raghav Chadha, Refuses Blanket Ban on Political Satire
Facts
Raghav Chadha filed a civil suit before the Delhi High Court against unidentified persons described as “Ashok Kumar/John Doe,” along with social-media intermediaries and government authorities.
The plaintiff alleged that unknown users had created and circulated AI-generated deepfake videos, morphed images and synthetic audio-visual content by copying his face, voice, mannerisms, speaking style and other aspects of his persona.
According to the plaint, the impugned online campaign began around April 2026 and contained material portraying the plaintiff negatively in relation to his political decisions.
The plaintiff sought protection of his personality and publicity rights and alleged that the content caused serious personal, political and reputational harm.
In his interim application under Order XXXIX Rules 1 and 2 of the Code of Civil Procedure, he sought, among other reliefs:
- restraint against the creation and circulation of AI-generated deepfakes and morphed content using his identity;
- removal of the identified URLs within 24 hours;
- removal of future or mirrored links upon direct intimation;
- disclosure of subscriber details, IP logs, telephone numbers and email addresses of the uploaders;
- blocking of fake accounts repeatedly publishing allegedly defamatory AI content; and
- directions to government authorities and service providers to block rogue websites carrying such material.
During the hearing, the plaintiff’s Senior Counsel stated that the plaintiff would not press the claim concerning personality rights and would confine the request for interim relief to defamation.
The Court therefore examined only the alleged defamatory content contained in Documents Nos. 1 to 52 filed with the plaint.
Issues
- Whether the impugned AI-generated and morphed content made out a prima facie case of defamation.
- Whether the plaintiff was entitled to a blanket interim injunction against all 52 items of online content.
- Whether political satire and criticism concerning the conduct of a public figure could be restrained before trial.
- Whether some of the impugned content crossed the boundary between protected satire and profane or malicious expression.
- Whether social-media intermediaries should be directed to remove specific URLs and disclose identifying information about the uploaders.
- How courts should balance a public figure’s dignity and reputation against freedom of speech and political criticism in cases involving artificial intelligence.
Plaintiff’s Arguments
The plaintiff initially argued that the unknown defendants had unlawfully misappropriated various facets of his persona, including his:
- name;
- image and visual likeness;
- voice;
- oratorical cadence and manner of speaking; and
- distinctive vocabulary.
He alleged that these features had been used through artificial intelligence, generative AI and machine-learning technologies to create deepfakes, voice-cloned audio and morphed videos for political, malicious or commercial purposes.
The plaintiff asserted that the synthetic content was derogatory, demeaning and deliberately designed to damage his political standing, public credibility and reputation.
He sought urgent takedown orders because the content could be rapidly copied, mirrored and circulated across several platforms.
However, when questioned by the Court, the plaintiff confined his interim case to defamation and did not press the claims relating to personality and publicity rights.
Respondents’ Position
The order does not record detailed substantive arguments on behalf of the intermediary platforms or government authorities concerning each of the 52 documents.
Meta Platforms and other defendants were represented during the proceedings.
The Court independently examined the impugned material and applied the heightened threshold governing pre-trial injunctions in defamation cases, particularly where the plaintiff is a politician and public figure.
Analysis of the Law
Requirements for an interim injunction
The Court reiterated that a temporary injunction is discretionary and may be granted only where the plaintiff establishes:
- a prima facie case;
- balance of convenience in his favour; and
- irreparable injury incapable of adequate monetary compensation.
Failure to establish any one of these requirements would disentitle the plaintiff to interim relief.
Personality rights distinguished from defamation
The Court observed that personality rights allow an individual to control unauthorised exploitation of his or her name, image, likeness and other identifying attributes.
Such rights generally arise where use of a person’s identity creates an impression of endorsement or commercially exploits the individual’s persona without consent.
However, since the plaintiff expressly declined to press the personality-rights claim at the interim stage, the Court did not decide those prayers and confined its analysis to defamation.
Pre-trial injunctions in defamation cases
The Court applied the principle that injunctions restraining publication before trial must be granted with exceptional caution.
A pre-trial restraint can seriously affect:
- the publisher’s freedom of speech;
- political discussion;
- the public’s right to know; and
- democratic debate.
An ex parte or interlocutory injunction should ordinarily not be granted unless the material is demonstrably malicious, palpably false or so clearly defamatory that any defence would inevitably fail at trial.
Higher threshold for public figures
The Court held that politicians and persons holding public office must tolerate a wider degree of criticism than private individuals.
Their public conduct, policy decisions, alliances and political actions are legitimate subjects of public discussion, satire and scrutiny.
Public figures are not deprived of dignity or reputation. However, they cannot seek restraint merely because a publication is unpleasant, annoying, exaggerated, inaccurate or critical.
The standard for obtaining an injunction is therefore higher where the impugned material concerns political activity or public conduct.
Artificial intelligence and constitutional rights
The Court stated that it did not endorse the use of AI-generated deepfakes or morphed media to harm an individual’s dignity.
Use of artificial intelligence may cross a constitutional boundary where it causes serious injury to the dignity and reputation protected under Article 21.
At the same time, AI is increasingly used to communicate opinions and satire, particularly in political discourse. Until comprehensive legislation regulates such content, courts must examine each grievance individually and balance:
- dignity and reputation; and
- freedom of speech and expression.
Precedent Analysis
Bonnard v. Perryman, [1891] 2 Ch 269
The Court relied on the traditional rule that interim injunctions in defamation cases should be granted sparingly.
Unless the alleged libel is clearly shown to be false and indefensible, courts should ordinarily avoid restraining publication before trial.
Bloomberg Television Production Services India Pvt. Ltd. v. Zee Entertainment Enterprises Ltd., 2024 SCC OnLine SC 426
The Supreme Court held that pre-trial injunctions against publication have severe consequences for freedom of speech and the public’s right to know.
Such orders should not be granted casually or ex parte unless the content is malicious or palpably false and the proposed defence would undoubtedly fail.
The High Court treated this as the controlling standard for examining the request to remove the impugned political content.
Kartar Singh v. State of Punjab, (1956) 1 SCC 692
The Supreme Court observed that persons holding public positions should not be overly sensitive to criticism.
Public officials must accept that attacks and adverse observations are an unpleasant but necessary consequence of public life.
R. Rajagopal v. State of Tamil Nadu, (1994) 6 SCC 632
The Supreme Court recognised that persons responsible for public administration must remain open to criticism.
Attempts to stifle criticism of government and public affairs may amount to objectionable political censorship.
Amish Devgan v. Union of India, (2021) 1 SCC 1
The Court referred to the need to balance freedom of speech with the constitutional values of dignity and reputation under Articles 14 and 21.
Freedom of speech remains essential to democracy, dissent and individual growth, but it does not render dignity and reputation irrelevant.
S. Charanjit Singh v. Aroon Purie, 1982 SCC OnLine Del 301
The Delhi High Court held that a Member of Parliament, being a public person, must tolerate searching criticism of his public conduct.
Naveen Jindal v. Zee Media Corporation Ltd., 2014 SCC OnLine Del 1369
The Court held that public figures should not obtain restraints merely because publications are inaccurate, distorted, annoying or offensive to their sensibilities.
Intervention may be justified where the allegations are grossly defamatory per se, but criticism during political activity and electioneering ordinarily receives wider protection.
Court’s Reasoning
The Court examined all 52 documents relied upon by the plaintiff.
It found that the majority concerned:
- changes in political affiliations;
- political alliances;
- governance;
- public policies; and
- decisions taken by the plaintiff in his political capacity.
Most of this content was characterised as political satire. The Court held that political decisions naturally attract both approval and criticism.
Satire, ridicule and humorous criticism concerning a politician’s public choices do not automatically become defamatory merely because they are unpleasant or damaging to the politician’s preferred public image.
The Court therefore refused to issue a blanket injunction against all the disputed material.
However, it found that Documents Nos. 2, 8, 9, 11, 25 and 40 contained explicit, profane and vulgar material. These six items went beyond harmless political satire.
The Court considered the balance of convenience to favour temporary restraint in relation to these specific documents.
Accordingly, it restrained the unidentified defendant from republishing the same content on other social-media platforms and directed the concerned intermediaries to take down the associated URLs.
The Court also ordered disclosure of Basic Subscriber Information and IP logs connected with those accounts so that the persons responsible could be identified.
No interim takedown was ordered in relation to the remaining documents because they primarily constituted satire or criticism relating to political conduct.
Conclusion
The Delhi High Court partly allowed Raghav Chadha’s interim application.
It refused to impose a blanket prohibition on AI-generated political satire or criticism concerning his political decisions, holding that public figures must tolerate a wider degree of scrutiny, ridicule and adverse comment.
However, it found that six items—Documents Nos. 2, 8, 9, 11, 25 and 40—contained explicit, profane and vulgar content falling outside protected political satire.
The Court directed Defendants Nos. 2 and 4 to:
- remove the URLs associated with the six documents within two weeks; and
- provide the plaintiff with the Basic Subscriber Information and IP logs of the relevant accounts within two weeks.
The matter was directed to be listed on 18 August 2026.
Case: Raghav Chadha v. Ashok Kumar/John Doe and Others
Court: High Court of Delhi at New Delhi
Case Number: I.A. 14417/2026 in CS(OS) 466/2026
Judge: Justice Subramonium Prasad
Date: 1 July 2026
Result: Interim application partly allowed; six vulgar and explicit AI-generated posts ordered to be removed, while blanket restraint against political satire was refused.
