Suyash Srivastava & Aarya Dubey are students at MNLU, Chhatrapati Sambhajinagar
INTRODUCTION
On May, 16, 2026, a satirical digital platform called the Cockroach Janta Party (henceforth, CJP or the Party) was launched after Chief Justice of India Surya Kant’s controversial courtroom remarks comparing certain unemployed youth to “cockroaches” and “parasites”. What began as an online joke quickly evolved into a nationwide virtual youth movement, amassing over 20 million followers on Instagram, eventually surpassing the Bharatiya Janta Party.
Soon, the movement gained momentum as it shifted its focus to the NEET UG 2026 paper leak, mobilising youth en masse to demand education reforms and the resignation of the Union Education Minister, Mr. Dharmendra Pradhan. It soon extended beyond the digital space, transforming its online support into a peaceful, on-ground protest, becoming India’s most visible youth movement in recent times.
However, as the protest grew, so did the divergence in its portrayal with some social media users getting to see its “good side”, while rest others, its “evil side” – both the sides controlled algorithmically. This raises a significant question – if different versions of the same reality are disseminated through algorithms controlled by foreign private corporations, has India’s democratic public sphere transformed into a privately controlled one?
This blog addresses the role of social media in shaping political narratives and argues that algorithmic control by foreign corporations poses a novel challenge to the democratic deliberations of a nation – a challenge that current legal frameworks are ill-equipped to address.
THE ALGORITHMIC CONTROL OVER OUR FEED
The defining feature of contemporary social media is that content visibility is no longer organic; it is algorithmically decided.
The Two-Feed Reality
The algorithmic mechanism of a social media platform is designed not to keep the users informed with the true, unbiased account of events but to maximise engagement. They use sophisticated machine learning models to learn user behaviour such as the accounts they follow, type of content they interact with, and the time they spend lingering on a post, and then curate a personalised feed based on their preferences. Algorithms that once subtly suggested movies, music, or humorous clips now shape the way we perceive the world.
In the context of the CJP protest, while one user engaging with content supportive of it came across videos of peaceful demonstrations and protestors getting assaulted by the “bearded policemen”, the feed of another user interacting with content opposing the movement was flooded with reels showing protestors getting violent with the police force, disrupting public order. Similar incidents are emerging worldwide. Historical events or current news can be portrayed in completely contradictory narratives based on users’ prior activities. This phenomenon of algorithmic curation limiting a user’s interaction with the content they are most likely to disagree with is known as Filter Bubble, a concept introduced by Eli Pariser in 2011. The result is a half-lived reality as both users encountering different feeds believe they have the complete picture and that the other has been misled. What they are unaware of is that their perceived reality was in fact, a constructed one.
Who Gets Seen?
Social media does not only decide what users consume but also determines the visibility of the content altogether. This is done by a phenomenon commonly called shadowbanning, i.e. reducing the content’s visibility without clearly informing the person who posted it as the content posted by the user stops appearing in recommendation feeds, hashtag searches, and explore pages, making it significantly harder to reach new audiences.
During the CJP protest, several Instagrammer protestors contended that their reels received unexpected drops in reach, delayed appearances on the Explore page, and lower engagement despite a large number of followers. One user wrote, “The shadow ban has arrived” and even contended that many protestors were facing suppression of content on Instagram. The Internet Freedom Foundation also urged Meta to explain the reasons behind such restrictions, noting that the company had not disclosed which specific content standards had triggered the actions. These incidents flag concerns about algorithmic transparency and accountability.
In Kaushal Kishore v. State of Uttar Pradesh (hereinafter referred to as “Kaushal Kishore case”), the Constitutional Bench of the Supreme Court (henceforth, SC), in a 4:1 majority, held that fundamental rights under Articles 19 and 21 can be maintained against private or non-State actors. Given this backdrop, if an intermediary silently downranks content without generating any warning or providing a reasonable explanation, does it not violate the guarantee of free speech and expression?
CONSTITUTIONAL VIOLATIONS BY THE ALGORITHMS
Democracy, the basic feature of the Constitution, requires citizens to deliberate based on shared facts (emphasis supplied). However, with phenomena like Filter Bubble and Shadowbanning, a lot of digital content is either filtered out or suppressed. This threatens the very possibility of collective political action, as a democracy cannot deliberate, protest, or hold anyone accountable if its citizens are not even responding to the same reality of events. It also prevents a large section of the public from accessing a diverse range of perspectives, including those which could have in fact challenged their partisan views, reinforcing their personal beliefs and making it harder to converse with alternate perspectives, thus creating a sense of naïve realism. What naïve realism produces, in effect, is a citizenry that believes their personalised feed reflects reality itself. They interpret disagreements as evidence that others are biased or irrational instead of realising that the algorithm has shown different people different pieces of reality.
Further, such filtration is violative of Article 19 (1)(a) of the Constitution, as the provision is not only restricted to a mere right of circulation of ideas but also includes the right to receive diverse sources of information, as highlighted by the SC in the seminal case of The Secretary, Ministry of Information v. Cricket Association of Bengal. This fragmentation of information into personalised silos leads to the circulation of repetitive ideas without meaningful opposition, thus intensifying artificial polarisation due to lack of exposure to contradictory ideas instrumental in any constructive dialogue. This results in the absence of open, adversarial exchange that the marketplace of ideas relies on. Although platforms may host vast user bases, algorithmic filtering creates new digital barriers to public discourse, especially for marginalised voices, whose perspectives may be further marginalised in these fragmented digital spaces, thus undermining their right to freedom of speech and expression guaranteed by the Constitution.
FOREIGN CORPORATIONS AS DEMOCRATIC INSTITUTIONS
Today, a major chunk of India’s democratic discourse is taking place digitally on platforms like Meta, Google and X, also known as the “Big Three” of social media.
It is to be noted that although these corporations are not State as per the verbatim of Article 12, they do have an obligation to protect free speech, as according to the Kaushal Kishore case, even private actors can be held responsible for the violation of Article 19 and 21. However, the intermediaries blatantly ignore this constitutional obligation and instead, they operate under a business model fundamentally misaligned with the epistemic and civic responsibilities historically associated with traditional journalism. This allows them to turn into “New Governors” of political speech, as also seen in the CJP episode, where content was shadowbanned and restricted without any explanation, leading to the privatisation of the public sphere.
This ultimately results in a transfer of authority to a foreign entity, answerable to no one while actively regulating political conversations, albeit algorithmically. It may therefore be argued that a part of India’s democratic sovereignty, has, in practice, been ceded to a private actor it never elected.
THE LEGAL VOID
The existing legal framework in India is not well-equipped to address the growing influence of algorithmic curation. While the SC’s decision in Shreya Singhal v. Union of India protected free speech by limiting governmental censorship to the grounds permitted under Article 19(2), it left a significant legal vacuum, primarily because it did not take into account the power of private platforms whose algorithms determine which voices are amplified and which remain unheard.
Although the Information Technology Act, 2000 read with the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 (henceforth, IT Rules, 2021) regulate content moderation by requiring intermediaries to remove unlawful content, they remain silent on algorithmic transparency, keeping the platforms free from any obligation to disclose the method of algorithmic functioning. This gives these platforms a free pass to influence the dissemination of information without corresponding legal accountability. While the legislature deliberating upon bringing amendments in the IT Rules, 2021 for the regulation of AI-generated content and deepfakes is indeed welcome, the major issue remains – Can a democracy afford to leave public discourse in the hands of opaque algorithms controlled by private corporations?
WAY FORWARD
A 2024 digital news report by Reuters Institute found that approximately 49 per cent of Indians rely on social media for news. This makes regulatory oversight necessary to hold intermediaries liable for their algorithmic designs that determine content visibility. The dependence of such a huge population on social media and the absence of transparency make addressing the issue vital for a democracy to thrive.
The Authors suggest an amendment to the IT Rules, 2021 mandating the social media intermediaries to disclose the parameters governing content recommendation and moderation. Such disclosure would bring transparency and enable users, researchers and regulators to understand how information is curated and distributed. Legislators can also study and draw lessons from the EU’s Digital Services Act, a legislation which requires such corporations to undergo systemic risk assessment, and independent algorithmic audits to enhance accountability for making the online environment safe and trustworthy.
Further, an independent statutory authority could be created for grievance redressal concerning shadowbanning or unjustified censorship. Presently, such grievances are largely addressed by internal grievance mechanisms of the intermediaries themselves, often lacking transparency and effective remedies. External supervision would provide users with a meaningful platform to challenge their grievances and would also strengthen public confidence in digital governance.
Finally, platforms themselves can adopt an inbuilt alert feature like “diversity nudges” that pops up when users are repeatedly engaging with similar content, raising a flag that they may be caught up in a filter bubble, suggesting them to view alternative perspectives. While the Authors acknowledge that this may not be a complete solution to algorithmic bias, it would still be a small step in widening informational diversity, thus contributing towards creating a more democratic digital space.
As algorithmic curation continues to shape democratic discourse, regulating it does not only remain a matter of digital governance but acquires constitutional importance, as a democracy cannot flourish when access to information is determined by opaque algorithms rather than informed citizenry.