On 17 November 2026, MediaNama reported that Instagram posts about Sonia Gandhi's memoir Belonging were restricted in India. The posts were published between 13 and 16 November. SFLC.in, a digital-rights group, says the restrictions are "being imposed by Meta's automated systems without specifying the reason or providing users with a clear way to challenge these decisions." The same report lists satire accounts, including Shamita Yadav's "The Ranting Gola" and Dr. Medusa, as restricted in India. Meta's side is not in the report, and the cause of each restriction is unconfirmed. Still, the episode shows a structural problem: a user cannot tell whether a platform, a court or a government ordered a restriction.
The case for fast, automated enforcement
The strongest argument for the current direction is a practical one. Deepfakes, non-consensual intimate imagery and incitement can do irreversible harm within hours, and platforms handle huge volumes that humans cannot review in time. Regulators who tighten deadlines are responding to real harms. In February 2026 the government amended the IT Rules so that compliance with lawful takedown directions fell from 36 hours to three, as LiveLaw's analysis reports. Free Press Journal reported that the rules take effect on 20 February 2026. The three-hour clock starts only after an official directive, which is a narrower trigger than a user complaint.
Why the design still fails speech
A three-hour window leaves little room for human review. When the cost of leaving content up is a legal penalty and the cost of removing it is invisible, platforms rationally over-remove, and automation makes that cheaper. MediaNama also reports an April 2026 claim, from an anonymous Meta source, that India had joined the countries where Meta automatically blocks flagged content. That claim is unconfirmed, so I treat it as a reported allegation rather than established fact.
The accounts affected here are political satirists and users discussing a book by a prominent opposition figure. Per MediaNama's account of SFLC's tracker, at least one satire account displays a notice that it is "not available in India, in compliance with IT rules." That wording points to a legal basis, but it names no order, no authority and no reason. It is not the same as a platform's own policy decision, and users cannot tell which they are facing. SFLC also records a former Gujarat corporator who says a Facebook post showing a selfie with the book was blocked by "government authorities." MediaNama does not verify that claim.
What Indian law already requires
India's blocking power is not meant to be opaque. Section 69A of the Information Technology Act, 2000 lets the Central Government direct an intermediary to block information only "for reasons to be recorded in writing," on listed grounds such as public order and the security of the State. The statutory text also provides that an intermediary that fails to comply faces up to seven years' imprisonment. That penalty explains why platforms over-comply.
The Supreme Court upheld the provision in Shreya Singhal v. Union of India (24 March 2015). It did so because of its safeguards. The judgment says reasons must be recorded in writing in a blocking order "so that they may be assailed in a writ petition under Article 226." It also describes blocking as possible "only by a reasoned order after complying with several procedural safeguards including a hearing to the originator and intermediary."
The test is simple. A restriction with no stated reason and no route to challenge it cannot be reviewed in court. If these restrictions come from government orders, that sits uneasily with Shreya Singhal. If they come from Meta's own classifiers, nothing in the Rules reviewed here gives the user a reasoned notice and a human appeal. I could not retrieve the Rules text from MeitY's site for this piece.
The courts are not yet providing clarity
The wider legal fight over executive takedown powers is stalled. On 22 July 2026 the Supreme Court, hearing transfer petitions filed by the Union government, stayed four High Court proceedings. They are X Corp's and DigiPub News Foundation's cases in the Karnataka High Court and Kunal Kamra's and Haresh Jagtiani's cases in the Bombay High Court. These challenge the Sahyog portal, as LiveLaw reported on 7 August. Until the Court decides, users have few places to test whether automated or government-triggered restrictions are lawful.
A proportionate fix
None of this requires dropping fast action against harmful content. Four measures would keep speed and add accountability:
- A statement of reasons. Every restriction should say whether it came from a government order, a court order or platform policy, and cite the rule or order.
- A real appeal. Users should get a human review within a fixed period, with the restriction lifted if the platform misses the deadline for non-urgent categories.
- Transparency reporting. Platforms should publish counts of automated restrictions by India-specific legal basis, so patterns like SFLC's become visible without volunteers collecting screenshots.
- Narrow automation. Automated blocking should be limited to categories with low contextual ambiguity, such as known intimate-image hashes. Political speech and satire need human judgement.
A book about a former party president, and satire about the government, are core political speech. If a restriction on them cannot be explained, it cannot be defended. The remedy is a reason and an appeal, and those preserve both the open internet and the legitimate goal of removing real harms.