India's Ministry of Electronics and Information Technology has asked Meta to re-examine how Facebook and Instagram decide what users see, warning that the platforms' recommendation systems may be amplifying narrow or one-sided narratives around sensitive public events. The advisory, dated October 10, sits at the intersection of content moderation policy and the technical architecture that governs modern social media feeds - a space where algorithmic choices increasingly carry legal and political consequences.
What the Advisory Actually Asks For
At its core, the directive targets the ranking and recommendation logic that determines which posts, videos and images rise to the top of a user's feed. The ministry wants Meta to ensure these systems do not disproportionately push a single viewpoint during periods of public sensitivity, and to add context - authenticity, date, location, circumstances - to visual content tied to unfolding events. It has also asked for firmer checks on material capable of inciting violence or distress, and for faster cooperation with law enforcement when such content surfaces.
These requests are framed as flowing from the due diligence obligations under the Information Technology Rules, 2021, which already require intermediaries to notify users about prohibited categories of content, including material that threatens public order.
Why Recommendation Algorithms Matter
Recommendation engines are not neutral conduits. They are optimisation systems, typically trained to maximise engagement signals such as time spent, shares and comments. Content that provokes strong emotional reactions - anger, fear, moral outrage - tends to perform well on these metrics, which can inadvertently reward divisive or decontextualised material over measured reporting. During fast-moving public events, this dynamic can mean that unverified clips or selectively edited images spread faster than corrections or clarifications ever could.
Asking a platform to "add context" sounds straightforward, but implementing it at scale is a genuine engineering and policy challenge. It requires systems capable of detecting when a photograph or video is being recirculated out of its original time or place, then surfacing reliable labels without becoming a bottleneck for ordinary content. Platforms including Meta have experimented with labelling and fact-check partnerships before, with mixed results.
A Contested Legal Foundation
The advisory has not gone unchallenged. Apar Gupta, founding director of the Internet Freedom Foundation, has called it a continuing censorship directive and questioned whether terms like "content without adequate context" or "heightened public-order sensitivity" correspond to any defined provision in the IT Act or the IT Rules. That critique points to a recurring tension in platform governance: advisories issued outside formal rule-making can carry real pressure on companies, even when their legal standing is ambiguous. Platforms operating in India must weigh compliance against the risk of overbroad takedowns that could chill legitimate speech.
The Broader Stakes
This episode fits a wider pattern of governments worldwide pressing large platforms to account for the societal effects of their ranking systems, from misinformation during elections to content moderation during unrest. For users, the underlying question is whether a handful of companies' engineering decisions should shape public understanding of contested events, and how much oversight is appropriate without tipping into state control of information flows. Meta's response - and whether it discloses any concrete changes to its systems - will likely shape how such advisories are treated going forward, both in India and in other jurisdictions watching closely.