The Ministry of Electronics and Information Technology has issued a formal advisory to Meta, directing the company to prevent its recommendation systems from disproportionately amplifying a narrow set of narratives and to monitor content capable of triggering public disorder or unlawful activity. The advisory, dated October 10 and reviewed by Scroll, arrives against the backdrop of protests in New Delhi demanding the resignation of Chief Election Commissioner Gyanesh Kumar over the special intensive revision of electoral rolls.
What the advisory actually demands
The ministry's communication asks Meta, which owns Facebook and Instagram, to supply "adequate context" around photographs, videos and graphics depicting ongoing public events. Its stated rationale is that content circulating without such framing "may contribute to public confusion, apprehension and disturbance of public order and social tranquility." The advisory does not name the Delhi protests directly, instead invoking the more clinical language of "periods of heightened public-order sensitivity" - a phrasing broad enough to cover almost any contentious public gathering.
Crucially, the ministry has asked Meta to review its "recommendation and content-ranking mechanisms" so that material deemed misleading, manipulated or stripped of context is not repeatedly pushed to users. It further instructs the company to cooperate promptly with law-enforcement agencies and to pay particular attention to posts containing unverified claims about emergency measures, police action or other official steps.
Why algorithmic amplification is the real battleground
Social media platforms do not display content in neutral chronological order. Ranking systems weigh engagement signals - shares, comments, watch time - to decide what appears in a user's feed, and these systems can inadvertently reward emotionally charged or divisive material because it holds attention longer. Asking a platform to avoid amplifying "a single or limited set of narratives" is, in effect, asking it to recalibrate the very logic that drives its business model. That is a far more consequential request than ordinary content moderation, since it touches the architecture of visibility itself rather than individual posts.
This also explains why such advisories are contentious. Unlike a takedown order aimed at a specific piece of content, instructions to adjust ranking behaviour are diffuse and difficult to verify from outside the company. There is no straightforward way for journalists, researchers or citizens to confirm whether or how a platform has altered its algorithm in response to government pressure.
Timing fuels suspicion of protest censorship
The advisory's issuance coincided with the Delhi Police detaining protesters and journalists ahead of an October 10 demonstration. Authorities also shut 45 metro stations, cancelled trains and suspended mobile internet in parts of central Delhi - measures organisers characterised as intimidation. Reports suggest roughly 7,000 protesters were detained, with allegations of assault during the crackdown.
The Internet Freedom Foundation has called for the advisory's withdrawal, arguing it risks enabling the "suppression of protest reporting." The digital rights group has also alleged that news outlets, journalists and activist accounts faced restrictions on Facebook and Instagram in the days following the advisory, and it has pressed Meta to disclose any content removals or visibility reductions linked to the government's intervention.
What is at stake for India's largest social media market
India is Facebook and Instagram's biggest market by user count, giving any government directive to Meta outsized reach. The tension here is structural: a government has legitimate interests in preventing misinformation from fuelling violence, but the same instruments - context labels, ranking adjustments, cooperation with law enforcement - can also be turned toward narrowing what citizens see about state conduct during protests. Without transparency on compliance, the public is left unable to distinguish genuine safety measures from quiet suppression of dissent.