Kaushik Raj*
While dismissing the lawyer’s repeated plea for senior designation on 15 May, CJI Surya Kant commented, “There are youngsters like cockroaches … Some of them become media, some of them become social media … and they start attacking everyone.” Without the context, that line became the name of a movement, the Cockroach Janta Party, and within days it gathered more than 20 million followers. By July, the meme merged with the NEET-UG paper leak scam, leading to almost twenty lakh students taking the paper again, with a string of student suicides. Then came 20 July, the CJP march towards the parliament, which ended up in tear gas and a lathi charge and further the resignation of the education minister. The CJI’s remark on a PIL concerning this action again got frothing on social media: “We aren’t interested in videos; we don’t have time to watch.” That wasn’t a PIL; not a single page has been filed. Which machinery decided how far and how fast this would travel, and who answers for that?
Virality is Not Manipulation, But It Proves the Stakes
No one outside the platforms involved can say with certainty what CJP actually was. It might be organic virality, a generation identifying with a random insult they felt and the real grievance, or maybe amplified by platforms that encourage outrage over anything, regardless of the reason. Both stories share the same set of facts, and there’s nothing on the public record that can tell us which is correct. But what the CJP demonstrates, no matter which side is right, is that India’s algorithmic structures can transform a sentence in days into a movement of 20 million followers who decide, unilaterally and without explanation, whether it keeps growing, gets throttled, or gets erased.
India’s legal system already has a language for this anxiety, but that runs in only one direction. A petition before the Supreme Court alleges that CJP’s rise is due to “commercial exploitation” and “algorithmic distortion” of judicial proceedings and seeks action against those responsible for this. What this does is treat the algorithm as a settled fact and prosecute it via the people that it carries, not as a matter of doubt over a system that decides to amplify what it does and reach without ever telling anyone or explaining it.
This isn’t limited to India. In 2025 Nepal, a government ban on social media collided with algorithmically driven protest content that helped fuel a youth-led protest movement into unrest that claimed the lives of more than seventy people, and a prime minister resigned. The same year in Los Angeles, researchers recorded algorithms promoting AI-generated content, such as a fake National Guard livestream that accumulated more than 900,000 views during a volatile protest. No matter the platform, there are different results but the same underlying condition: infrastructure that has the impact and power of deciding which grievances scale, and no one has a duty to explain how. A platform’s amplification or throttling decisions shape which movements live and which die, and no law mandates that decision to be disclosed to the person affected, to a court, or to the public. The only accountability mechanism on offer is retrospective and punitive, aimed at the people the algorithm carried, never at the algorithm itself.
A Sentence the Court already Wrote
Buried in the 2017 Puttaswamy judgment, there’s a line that has never been used for what it now describes. While discussing the protection of privacy, Justice Chandrachud held that mental integrity entitles a person to “the freedom of thought, the freedom to believe in what is right, and the freedom of self-determination”.[1]. The court went further, stating that one of the nine aspects of the right that it had just recognised was intellectual privacy, privacy of thought and development of opinion. Justice Nariman picked up the same idea from the Preamble, which says that the state will ensure “liberty of thought, expression, belief, faith and worship”[2] and termed Article 21 as the “most fully” article embodying liberty.
Eight years later, that one thing is yet to be described in that language. Puttaswamy safeguarded the right to free thoughts, to keep one’s beliefs private and undisturbed, from intrusion by the state or a neighbour who looks in. There was no need for it to think about another threat: not the threat of it being imposed on you, but the threat of it being constantly shaped by a feed and influencing your thinking before you even have a chance to think about it. That’s not the same, right? It’s the one Article 21 has already identified, waiting for a case in which the shaping entity is not a person but an architecture that is accountable to nobody.
This does not need to create a new right. The Court has already ruled that a clean environment, public health and education are covered by the principle of Article 21, and none of these is expressly stated in the text of the article. The method is no big jump to cognitive liberty. It’s the way of arriving at a sentence the court wrote in 2017 and never had reason to use.
A Regulatory Patchwork with a Hole in the Middle
The problem is clearly spelt out in the petition filed before the Supreme Court in Vishal Tiwari v. Union of India for the ‘Rs. 370 biryani’ incident: Existing legal mechanisms are mostly reactive and get activated only when false information has reached irreversible virality. It isn’t wrong. From the DPDP Rules, 2025, it is required that large platforms conduct algorithmic due diligence to ensure that their systems do not pose a risk to users’ data rights. New rules under the IT Amendment Rules 2026 mandate labelling of AI-generated content and retention of the source, with takedown windows as short as 2 hours for the most harmful deepfakes.
Neither side gives any consideration to the current scenario. One is about the way in which personal data is dealt with. The other is the disclosure of AI content. Nothing in either requires a platform to say. Nothing in either requires a platform to reveal how a particular account got 20 million followers in a matter of days and why it was later revoked, or how a hashtag was able to run as long as it did. India regulates the data. India regulates the fakes. It has not been required to make a regulation to control a decision.
Auditing the Algorithm, Not Policing the Speech
None of this argues for a power to take down, block or punish content. It advocates the right to information about decision-making. Platforms must be held accountable, not case by case, to disclose the criteria on which they base their decision to amplify or throttle anything. When a disclosure reaches a defined threshold, then it should be expanded into an impact assessment, a detailed analysis of how those criteria played out in practice. All users should be given the choice to opt out and return to a normal feed that just runs sequentially.
Give that power to a state instead of an auditor, and it gets used the way it already has. CJP’s own account was withheld in India because of existing powers of the government. In the other PIL filed before the Delhi High Court requesting an NIA probe into the same protests, the court, taking only a technical view, refused to grant the permission, stating that this was the job of the Union government and not the court. Nothing about the algorithm was ever in question. Only those who get investigated for what it carried.
The only difference between an auditor and a censor is that an auditor must explain a decision while the censor just makes it disappear.
The letter which raised the police action on 20 July never became a case. It was one representation, unread in any record, answered with a line that later needed its own correction. Regardless of what happens to this petition or the one calling for an NIA probe or the CJP itself, the question remains: who explains how a sentence uttered in an unrelated hearing became a movement among the 20 million people and who explains how it was throttled? The preamble has carried “liberty of thought” since 1949, waiting for Article 21 to ask what that means when the thing doing the shaping is a feed, not a person. It is still waiting.
* The author is a second-year Law student at National University of Study and Research in Law (NUSRL), Ranchi. The author can be contacted at kaushik.raj@nusrlranchi.ac.in.
This blog reflects the personal views of the author and does not necessarily represent the views of The Policy Chronicle.
[1] K.S. Puttaswamy (Privacy-9J.) v. Union of India, (2017) 10 SCC 1
[2] India Const. pmbl.