MUMBAI, Aug. 6, 2026 - The Bombay High Court has directed major online platforms to remove social-media posts and videos identified by Union Road Transport Minister Nitin Gadkari as fake, obscene and defamatory, including AI-generated or manipulated material linking him personally to the controversy over E20 ethanol-blended petrol.
The court granted urgent interim relief on Wednesday in proceedings involving Meta, X and Google. Reports of the hearing say the judge described the material placed before the court as vile, abusive and prima facie defamatory, and questioned why platforms did not have an effective system to respond quickly when such posts were brought to their attention. The platforms were told to take down the specified content without waiting for the larger lawsuit to be finally decided.
The scope of the order needs careful explanation. Gadkari's case was not simply that people criticised the government's decision to expand the use of petrol blended with 20 percent ethanol. Public policy can be questioned, and strong disagreement with a minister is ordinarily part of democratic debate. His complaint concerned posts that allegedly used manipulated videos, false images or fabricated statements to portray him as corrupt or secretly benefiting from the policy. The legal issue is therefore the line between protected criticism and knowingly false factual allegations that damage reputation.
Deepfakes make that line harder to police. A conventional false post may be rebutted with documents or a correction. A convincing synthetic video appears to show a person saying or doing something that never happened, and it can spread across several services before the subject even knows it exists. Copies may be cropped, translated or reposted under new accounts. By the time a platform reviews the original URL, dozens of near-identical versions may already be circulating.
Courts increasingly face requests for dynamic injunctions that cover not only a specific post but also substantially identical copies. Such orders can be useful when content is clearly fraudulent, yet they also require safeguards. If similarity tools are too broad, they may remove lawful commentary, parody, reporting or evidence documenting the original misinformation. The safest approach is to define the prohibited material precisely, require prompt review of notified copies and preserve a route for users or publishers to challenge mistaken removal.
India's wider regulatory framework is still adapting to generative AI. Existing rules on defamation, impersonation, privacy, intermediary obligations and election conduct can apply to synthetic content, but enforcement is divided among courts, police, regulators and platform processes. A durable response will need more than celebrity lawsuits. Platforms should label provenance where reliable, retain upload records for lawful investigation, make complaints easy to track and disclose how often synthetic-media notices lead to removal. Government rules must be narrowly written and independently reviewable so that a tool created to stop impersonation is not used to erase satire or uncomfortable reporting.
The case also highlights the difference between platform knowledge and general monitoring. Online services process an enormous volume of speech and are not expected to know automatically that every disputed claim is false. Once they receive a properly documented court order or notice identifying particular unlawful content, however, delay becomes harder to justify. The High Court's questions focused on whether the companies' systems were capable of moving with the speed required when reputational harm was continuing in real time.
Because the order is interim, it does not finally decide all contested facts, damages or platform liability. The defendants may place their legal positions before the court, and the judge can later modify the relief. An interim injunction is designed to prevent continuing harm while the suit proceeds. Readers should therefore avoid describing the hearing as a final judgment establishing every allegation made by Gadkari.
The E20 debate itself should remain open to evidence. Motorists, vehicle makers, fuel retailers, farmers, environmental researchers and economists may raise legitimate questions about mileage, compatibility, feedstock, emissions, food-versus-fuel trade-offs and pricing. Removing a fabricated video does not answer those questions. In fact, separating false personal allegations from policy analysis can improve the debate by forcing both supporters and critics to rely on verifiable data.
For newsrooms, the episode offers a practical verification checklist. Before embedding a sensational clip, editors should find the earliest upload, examine audio and visual discontinuities, seek the full recording, compare it with official transcripts and contact the person depicted. A visible label saying that a clip is unverified is not always enough if the headline repeats the false claim. News organisations should also retain evidence of the manipulation while avoiding unnecessary republication that increases its reach.
The High Court's intervention is significant because it treats synthetic media as a concrete legal problem rather than a futuristic threat. Still, the ruling should not be reported as judicial censorship of the E20 controversy. The accurate description is narrower: the court ordered the immediate removal of identified content it considered prima facie defamatory, while the underlying civil action and the public debate over ethanol policy continue.
