When 'Information' Swallowed Platform: Section 69A After Telegram FZ LLC

For eleven years, anyone reading Section 69A of the Information Technology Act, 2000 could take one thing for granted: the provision blocked content, not the platform carrying it. The Supreme Court upheld Section 69A in Shreya Singhal v. Union of India (paragraph 109) precisely because it was narrowly drawn. Blocking could rest only on the grounds in Article 19(2). Reasons had to be recorded in writing. The intermediary got a hearing before the Rule 7 Committee before any order was confirmed. The provision survived because it was aimed at identifiable content moving through a network the government did not otherwise touch.

The Delhi High Court's judgment in Telegram FZ LLC & Anr. v. Union of India & Ors. (W.P.(C) 8259/2026), delivered by Justice Tejas Karia on 19 June 2026, changes what the provision can reach. On 16 June, MeitY issued an interim order under Section 69A blocking Telegram across India until 22 June and separately directing Telegram to disable its message-editing feature until 30 June, all timed to the NEET-UG re-examination of 21 June and the paper-leak allegations that preceded it. The Rule 7 Committee heard Telegram on 17 June and the order was confirmed. Telegram's petition against both directions failed, and it failed on a ground much larger than the six-day block that provoked it. The Court held that “information” under Section 2(1)(v) of the IT Act includes codes, computer programmes, software and databases, and that an application is nothing more than a compilation of exactly those things. Once a platform is itself “information”, blocking the platform differs from blocking a URL only in scale, not in kind. That is a genuinely new proposition, and it deserves to be treated as one rather than folded quietly into the settled reading of a fifteen-year-old provision.

What Made Section 69A Narrow

Shreya Singhal did not uphold Section 69A because blocking powers are harmless. It upheld a specific design: satisfaction confined to Article 19(2) grounds, reasons in writing, and the procedure of the 2009 Blocking Rules, under which the Designated Officer's request goes to a Rule 7 Committee, the intermediary is heard, and only then does a final order issue.

Notice what those safeguards regulate. They regulate why something is blocked and how that decision is tested. They say nothing about how much is blocked. The assumption doing the real work underneath them was never argued, because nobody needed to argue it: the “information” being blocked was something a user posted or hosted, while the platform carrying it stayed available for everything else on it. Every application of Section 69A since has rested on that assumption, whether the order blocked one post or, as has become routine, thousands of URLs in a single tranche. The volume grew. The kind of thing being blocked did not.

What the Telegram Court Did Instead

The Court split the case into two issues: whether the order suffered from non-application of mind, and whether the blocking was proportionate. On the second, it expressly framed two prongs at paragraph 32: first, whether the government was legally empowered to block public access to Telegram at all, and only then, whether the exercise of that power was disproportionate. The order of those two prongs is where the judgment's real significance lies.

On the power question, the government's position was that Section 69A never actually drew the line the petitioners assumed. Section 2(1)(v) defines “information” to include data, messages, text, images, sound, voice, codes, computer programmes, software and databases. An application, on any ordinary understanding, is a compilation of code, databases and the interfaces that connect them. If code and databases are information, an application is information. And Section 69A permits the blocking of any information generated, transmitted, received, stored or hosted in any computer resource.

The Court accepted this reading in full. At paragraph 35, it held that confining “information” to individual user accounts, channels, images, posts, files or messages would be an unduly narrow construction that may render the provision otiose. The legislature, it said, intended a broad and technologically neutral meaning. At paragraph 37, it went further: there was no reason to exclude an application or platform from the definition, since an application is a computer programme or software designed to perform functions for an end user, and the definitions of “computer resource” and “computer” in Sections 2(1)(k) and 2(1)(i) confirm that information generated, transmitted, stored or hosted through such software-based infrastructure falls within Section 69A.

Two things follow. First, the government was empowered to block Telegram in its entirety, not merely the 1,300 or so URLs it had directed Telegram to take down in the weeks before the platform-wide order. Second, and more importantly, this is not an extension of Section 69A by analogy. It is a holding that the provision always covered whole platforms, and that years of URL-by-URL practice were simply the government choosing not to use the full power it had. Nothing in this reasoning is specific to Telegram's technical quirks. It is generic statutory interpretation, and it applies to any application a court is prepared to call a compilation of code and databases. That means every application.

Duration Is Not Scope

Having found the power, the Court tested its exercise against Anuradha Bhasin v. Union of India. The petitioners argued that a platform-wide block failed the requirement that only the least restrictive measure be adopted. The Court quoted paragraphs 78 to 80 of Anuradha Bhasin, reproduced at paragraph 38 of its own judgment, including the requirement that a restriction's degree and scope, both territorially and temporally, must stand in relation to what is actually necessary to combat an emergent situation.

Applying that test, the Court found two things. Channel-by-channel takedowns had failed, because banned channels reappeared through mirror channels, backup channels, bots and rotated handles. And the measures were short: the platform block ran only until 22 June, a day after the examination, while the message-editing suspension ran until 30 June. At paragraph 46, the Court said the limited temporal scope of these measures demonstrated that they were narrowly tailored, and on that basis held the requirements of proportionality fulfilled.

But look at what each finding actually answers. The temporal finding answers duration. The findings on failed takedowns answer whether anything narrower would have worked against this platform. Neither answers the question the power holding created: how large a category of executive action has just been declared lawful. A six-day window genuinely limits how long a harm lasts. It says nothing about the size of the power. Brevity and narrowness are independent variables, and they moved together here only because MeitY chose not to run the block longer than the exam required. The editing-feature direction makes the point sharper, not weaker: it ran nine days past the exam and ordered a change to how the software functions, which is even further from blocking a piece of content than the block itself was.

The Strongest Objection, and Why It Does Not Hold

The fair response is that the Court did not hand out platform-blocking power in the abstract. Its holding rested on findings specific to Telegram: cloud-based storage, a large bot ecosystem, username-based operation that conceals identifiers, and message-editing with timestamp retention, which could make a paper appear to have leaked before an exam when the edit came after it. The Rule 7 Committee found that real-time segregation of legitimate communications from illicit content on the platform was practically unfeasible, and channel-specific enforcement had already been tried and had failed. The record was severe. The Rajasthan Police Special Operations Group had documented Telegram-based circulation networks in its investigation into the cancelled 3 May examination. The Ahmedabad cyber crime branch had arrested an inter-State gang running eight Telegram channels, with about Rs 1.5 crore in documented fraudulent transactions. And Telegram's own chief executive had posted publicly on 16 June that the platform had removed hundreds of channels sharing leaked exam material in India, a statement the Committee treated as confirming both the scale of the misuse and the platform's reactive, after-the-fact posture. On this reading, the judgment stands for something narrow: a platform whose own design defeats granular enforcement, on a record this severe, can be blocked wholesale. A fact-bound holding, not a doctrinal one.

That is largely correct about what the Court required. But look at where those findings sit in the judgment's own structure. They appear at paragraphs 42 to 45, inside the proportionality analysis, after the Court had already held at paragraph 37, on pure statutory construction, that Section 69A reaches applications as such. The architecture answers whether blocking Telegram was the least restrictive way to achieve the government's object. It was never asked to answer whether Section 69A can reach a platform at all. That question was settled first, under the first prong the Court itself framed at paragraph 32, before any inquiry into bots or mirror channels began.

Because the architectural showing operates as a proportionality threshold rather than a doctrinal limit, it travels to any platform whose design produces a comparable record, and building that record is a matter of degree, not of kind. WhatsApp's end-to-end encryption already frustrates the granular, content-level takedown Section 69A traditionally assumed, in terms a regulator could adapt from the Telegram order without much strain. X's decentralised moderation and rapid account re-creation after suspensions map onto the audience-migration finding this judgment relied on. Signal is more resistant to content-level enforcement than Telegram, not less, which on the Court's own logic makes it a stronger candidate for a whole-platform block, not a weaker one. None of this requires a future court to extend Telegram FZ LLC. It requires only a future government to assemble the same kind of dossier MeitY assembled here, built from a platform's own advertised design features. That is a considerably lower bar than showing that specific unlawful content exists and cannot otherwise be reached.

What Is Now Available, Regardless of What Was Intended

None of this requires believing MeitY set out to create sweeping authority, or that Justice Karia intended a template for blocking WhatsApp or X. The order responded to a genuine emergency: a national medical entrance examination for 2.2 million candidates, channels operating openly under names like “PAPER LEAKED NEET” and “Re-NEET 2026” that demanded money from candidates and their families, and a documented pattern of channels reappearing within hours of takedown. Tested against those facts, the specific measure was defensible.

The difficulty is that judicial reasoning does not expire with the emergency that produced it. The construction of “information” in Section 2(1)(v) to include an application's codebase and databases is now available to every future Designated Officer considering every future platform. It carries no sunset clause tied to examinations, six-day windows, or Telegram. It carries only an evidentiary threshold, and that threshold is built from the very features that make a platform valuable to its users in the first place: encryption, decentralisation, resistance to centralised takedown.

Shreya Singhal drew its line at content because that was the line Parliament drew and the 2009 Rules were built to police. Telegram FZ LLC has not repealed that line by amendment or read it down by express holding. It has found, as a matter of statutory construction now available for citation in every future blocking dispute, that the line was never really there. The next platform this reasoning reaches will not need a paper leak or an examination to justify it. It will need only a Designated Officer, a Rule 7 Committee willing to make the findings this one made, and a platform good enough at protecting its users that a court can be persuaded nothing narrower would do.

 Author is an advocate practicing at Rajasthan High Court, Jodhpur. Views are personal.