India's electronics and information technology ministry has formally asked Meta to re-examine the systems that decide what Facebook and Instagram users see, warning that the company's algorithms risk funnelling people toward a single narrative instead of a balanced range of perspectives. The advisory, dated October 10, also demands tighter safeguards against content capable of inciting violence or causing public distress, placing fresh scrutiny on the mechanics of algorithmic recommendation at a moment of heightened sensitivity around public order.
What the Advisory Actually Demands
The Ministry of Electronics and Information Technology (MeitY) has outlined several concrete expectations rather than vague appeals for caution. Meta has been told to provide users with context around photographs, videos and graphics tied to public events, so that authenticity, date, location and the circumstances of creation or sharing are not left ambiguous. The ministry has also asked the company to review its recommendation systems and ranking mechanisms to curb the spread of misleading or manipulated material, and to cooperate promptly with law enforcement and other government bodies when required. The advisory pointed specifically to content linked to ongoing situations involving public order, noting that such material was being promoted with unusual prominence on Meta's platforms.
The Legal Framework Behind the Request
MeitY has grounded its demand in the Information Technology Act, 2000, and the due diligence obligations set out under the Information Technology Rules, 2021. Those rules require intermediaries to inform users about categories of prohibited content, including material that could incite violence or disturb public order. This is the legal scaffolding the government leans on whenever it issues advisories of this kind, and it is why the framing matters: an advisory is not a court order or a formal regulation, but it carries the implicit threat of intermediary liability if platforms are seen as non-compliant with due diligence norms.
Why Recommendation Systems Are the Real Battleground
Content recommendation engines are the invisible architecture of every major social platform. They decide, through engagement signals and behavioural data, which posts rise to the top of a feed and which fade into obscurity. Critics of these systems have long argued that optimising purely for engagement tends to reward emotionally charged or divisive material, because such content keeps people scrolling. Asking a company to ensure its algorithm does not disproportionately expose users to one narrative is, in effect, asking it to recalibrate a system built around attention rather than balance - a far more complex technical and commercial undertaking than removing a single piece of flagged content.
Pushback From Digital Rights Advocates
The advisory has not gone unchallenged. Apar Gupta, founding director of the Internet Freedom Foundation, has called it a continuing censorship directive and questioned its legal footing. He pointed out that terms used in the advisory, such as content lacking "adequate context" or periods of "heightened public-order sensitivity," are not tied to any specific clause of the IT Act or the IT Rules. That ambiguity, in his view, leaves platforms guessing at compliance and gives the government considerable informal leverage over moderation decisions without the accountability that comes with formal rule-making.
The episode reflects a broader tension playing out across democracies: governments want platforms to curb harm and misinformation, but the tools available to do so - recommendation tweaks, context labels, takedown cooperation - sit uneasily between technical design choices and speech regulation. How Meta responds, and how transparently, will shape the next round of this debate.