MeitY Asks Meta To Review Content Recommendation Algorithms

MeitY asked Meta to change how Facebook and Instagram recommend and present content about ongoing public events. The ministry wants users to receive diverse, relevant, and credible information rather than disproportionate exposure to one narrative. It also wants clearer context when a photograph, video, graphic, or statement may have unclear authenticity, date, location, or circumstances.

The advisory specifically highlights old or selectively edited visuals presented as current, unverified claims about emergencies or police action, and law-and-order material lacking context. Meta must review its ranking systems, identify content gaining significant reach, and rapidly escalate material that could cause confusion, apprehension, violence, or disturbances. It must also cooperate with authorised agencies and inform MeitY about action taken.

The measures were requested with immediate effect under the Information Technology Act, 2000, and IT Rules, 2021. The advisory also says enforcement must respect lawful expression and legitimate public discourse. Digital-rights advocates nevertheless criticised it as a potentially broad censorship directive.

A recommendation algorithm is a platform’s automated system for selecting, ordering, and displaying content to users. It helps decide which posts, videos, images, or statements appear prominently, are suggested next, or are repeatedly recommended. This matters because users may encounter recommended material more often than content they actively search for.

In general, such systems use signals connected to a user’s activity, interests, and interactions, along with characteristics of the content and its popularity. The article does not list Meta’s specific signals. It reports that government agencies believed Meta’s systems appeared to favour certain types of content over others, and MeitY therefore requested a review of recommendation and content-ranking mechanisms.

The advisory seeks to prevent these systems from disproportionately amplifying misleading, manipulated, or out-of-context material. It also asks Meta to provide a diversity of relevant and credible information. This approach recognises that ranking choices can influence which narratives become most visible during sensitive public events.

During a live public event, users often see a fast-moving stream of images, videos, statements, and graphics. If an old visual is presented as current, or an edited clip is shown without its original circumstances, viewers may mistake it for evidence from the event underway. The platform’s recommendation system can make that mistake more influential by displaying the material prominently or suggesting it repeatedly.

MeitY specifically mentioned old or selectively edited visuals presented as current events, unverified claims about emergencies and police action, and audio-visual material lacking law-and-order context. The key mechanism is amplification: ranking systems can give selected material wider and more repeated exposure than ordinary posts receive. Missing captions or background can make verification harder.

The advisory asks Meta to provide adequate context about authenticity, date, location, and circumstances. It also seeks stronger systems to identify, review, and escalate content that gains significant reach and could create public confusion, apprehension, violence, or disorder.

A recommendation system can reach potentially very large audiences because it operates across Meta’s platforms, including Facebook and Instagram. The article does not provide a specific number of people affected. Its concern is scale: content that is prominently displayed or repeatedly recommended can travel far beyond the people who first posted or viewed it.

Repeated recommendations can make one narrative occupy more users’ attention than alternatives. Each additional placement creates another opportunity for exposure, while competing explanations may receive less visibility. During a public event, this can make an unverified claim or a selective account appear widely accepted, even when its authenticity, date, location, or circumstances are unclear. The article describes this as disproportionate exposure.

MeitY wants recommendation systems to provide appropriate diversity of relevant and credible information. It also wants Meta to review material that gains significant reach and could cause confusion, apprehension, violence, or public-order disturbances. The requested changes focus on amplification, not merely individual posts.

The Information Technology Act, 2000, and the Information Technology Rules, 2021, form the legal framework MeitY invoked in its advisory to Meta. In the article’s description, these measures require online intermediaries to exercise due diligence and inform users about prohibited content. The framework is therefore connected to platform responsibilities for handling content and communicating rules to users.

The advisory also called for timely cooperation with law-enforcement agencies and other authorised government bodies. MeitY used this framework while asking Meta to review recommendation and ranking mechanisms, strengthen content checks, and rapidly identify material that could cause disorder or incite violence. Meta was asked to implement the measures immediately and report what action it took.

The article also records an important legal dispute. Internet Freedom Foundation founder-director Apar Gupta argued that the advisory was not legally binding and questioned its legal basis. He said phrases such as “without adequate context” were not tied to specific provisions. The advisory itself says actions must comply with applicable laws and respect lawful expression.

If unverified claims about protests, emergencies, or police action are widely recommended, users may form an inaccurate picture of events. The article warns that material circulated without adequate context or amplified disproportionately may contribute to public confusion, apprehension, and disturbance of public order and social tranquillity. It also identifies content that could incite violence or other unlawful activity.

For example, an old or selectively edited visual presented as current could be interpreted as evidence of an emergency or police action. Repeated recommendations could give the claim significant reach before its date, location, authenticity, or circumstances are clear. The advisory therefore asks Meta to identify, review, and escalate content that gains major reach through recommendation systems.

The requested response is not limited to removing individual posts. MeitY wants stronger ranking checks, added context, and rapid review of potentially harmful material. It also asks Meta to cooperate with authorised agencies. At the same time, the advisory says any action must respect lawful expression and legitimate public discourse.

Regulating recommendations is difficult because platforms must address possible harm without treating every disputed or unpopular view as dangerous. The government’s advisory focuses on content that may cause public confusion, violence, or public-order disturbances. It also asks for context and a diversity of credible information. These goals involve judgement about accuracy, relevance, urgency, and likely effects.

The article gives a concrete tension. MeitY wants Meta to review ranking systems and rapidly escalate content gaining significant reach. The advisory says measures must comply with law and respect lawful expression and legitimate public discourse. Yet Internet Freedom Foundation founder-director Apar Gupta called it a “continuing censorship directive” and argued that it was not legally binding.

Gupta also questioned phrases such as “without adequate context” and “heightened public-order sensitivity,” saying they were not tied to specific legal provisions. The forward issue is transparency and accountability: he urged Meta to disclose every takedown and downranking carried out in response. The article does not state whether Meta adopted that proposal.