There are moments in governance when a procedural change reveals a deeper ideological, even philosophical, shift. The Union government’s decision to compress the takedown window for “unlawful” online content to two or three hours is one such moment.
On the face of it, the amendment to the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021, appears technical. In substance, it is illiberal, predatory and coercive. Social media platforms now have just two hours to remove “non-consensual intimate imagery” and three hours to take down all other categories of “unlawful” content. All this is reduced from the earlier 24–36 hour window.
Sources among industry executives point out that this is one of the shortest mandated compliance timelines anywhere in the world. That fact alone signals that this is not incremental change. It is a regulatory acceleration of an extraordinary order. This is not merely a change in hours. It is a reordering of power and structures.
Three Hours to Judge the Law
What constitutes “unlawful” content in India is not a narrowly defined category. The legal architecture governing speech spans the Indian Penal Code, the IT Act, anti-terror laws, defamation statutes, sedition provisions (however rephrased), public order doctrines, and an array of special enactments. Political speech, satire, parody, investigative reportage, dissent, criticism of public officials—each has, at various points, faced legal scrutiny.
Under the amended rules, platforms must now make determinations about such content in less time than it takes to conduct a board meeting or argue a bail application.
Failure to comply risks the loss of safe harbour protections under Section 79 of the IT Act—the shield that prevents intermediaries from being treated as publishers of user-generated content. Strip that shield away, and every post becomes a potential liability; every delay, prosecutable exposure.
The architecture is unmistakable: define illegality broadly, compress the compliance window drastically, and attach existential risk to non-compliance. The State does not need to issue explicit censorship directives. The incentive structure ensures the outcome. Faced with legal uncertainty and microscopic timelines, platforms will over-comply. As the media in general, especially sections of the print and television broadcast media, have shown in recent years, the normal response under threat is deletion. When in doubt, remove. Due process shrinks to fit the clock.
Consultation as Afterthought
Technology companies have insisted they were not formally consulted before the amendments were notified. Government officials counter that there were “sufficient discussions” and that citizens demanded quicker redress against harmful content.
The pattern is familiar. Major digital policy decisions—from the original 2021 Rules to data governance frameworks—have often been unveiled after limited or opaque stakeholder engagement. Consultation, where it occurs, tends to be selective and retrospective rather than foundational. A former Kerala Chief Minister used to say that velocity should define governance style. The IT ministry in Modi 3.0 seems to have quantum-leaped to this style.
Determining illegality frequently requires contextual assessment. Law enforcement communications are not always precise. Moderation at scale already produces documented errors—false positives, wrongful removals, uneven enforcement. Compressing the response time does not eliminate ambiguity; it intensifies it. Three hours does not allow for layered review, proportionality, or meaningful appeal. It shifts adjudication from courts to compliance desks operating under legal threat.
The Stamp of Geopolitics and Trade Deals
The timing of the amendment is geopolitically interesting. Only days earlier, India and the United States announced a framework for an interim trade arrangement that spoke of reducing “discriminatory or burdensome practices” in digital trade and crafting mutually beneficial digital rules.
India’s digital ecosystem, however, is dominated by US-headquartered corporations—Meta, Google, YouTube, X, OpenAI, Adobe and others. While the rules formally apply to Indian platforms as well, the compliance burden will weigh heavily on American firms with vast user bases.
Washington has not historically remained silent when regulatory pressure is applied to its technology giants abroad. The United States has criticised European Union enforcement actions against American tech firms, portraying them as punitive or protectionist.
If India’s compliance regime is perceived as disproportionately onerous, the matter may migrate from domestic regulation to diplomatic friction. Digital sovereignty is a legitimate aspiration. But sovereignty exercised amid trade negotiations introduces another layer of strategic calculation.
AI Ambition, AI Anxiety
The amendments also reshape India’s regulatory posture toward artificial intelligence-generated content. An earlier draft had proposed that AI labels occupy at least 10% of the content space—a visually intrusive requirement that triggered strong resistance from media practitioners. The final version retreats from the numerical threshold but insists that labels be “prominently visible”.
That concession is evidently cosmetic. The deeper obligations are expansive. Platforms must act “expeditiously” upon awareness of synthetic content that violates the law. They are required to deploy “reasonable” technical measures to prevent AI-driven misrepresentation of real-world events or identities. Users must declare when content is synthetically generated and verify and label it prominently.
The vocabulary—“reasonable,” “appropriate,” “expeditious”—is elastic enough to fit in with the Rashtriya Swayamsevak Sangh (RSS)-led Sangh Parivar’s time-tested Multi-Speak strategy. Elastic language paired with compressed timelines invites expansive interpretation. Here lies the central contradiction.
India proclaims its ambition to become a global AI powerhouse. It hosts summits, announces innovation missions, speaks of indigenous large language models and digital public infrastructure as exportable templates. The India-AI Impact Summit, concluding on February 20—the very day the amended rules take effect—celebrates technological leadership.
Yet alongside this ambition runs a regulatory reflex of control. The same State that encourages AI innovation now mandates real-time policing of AI output under severe liability risk. An AI superpower governed by a stopwatch.
Innovation thrives in environments of regulatory clarity and predictable enforcement. Uncertainty combined with punitive timelines may not stifle technology overnight—but it alters incentives. Indeed, the government and its agencies have a point when they assert that digital harm travels at great speed and that deepfakes, disinformation campaigns, non-consensual imagery, and coordinated hate can damage reputations and destabilise communities within minutes.
The harm is real. But urgency cannot replace standards. A democratic framework traditionally balances harm prevention with procedural safeguards. The three-hour rule compresses that balance. It relocates power from judicial oversight to executive notification and corporate compliance units. This is not just about faster grievance redress. It is about reconfiguring the site of decision-making. In other words, this is also an executive assertion that censorship need not be declared; it can be made structurally rational.
The Legal Challenges Ahead
There is little doubt that the new stipulation would be challenged in courts. By all indications, industry associations are getting ready to challenge the proportionality of the timelines. Civil liberties groups are also likely to test the constitutional validity of compressed due process. Trade negotiators may quietly raise questions.
Yet beyond the courtroom and the negotiating table lies a more fundamental inquiry. Can a democracy preserve a vibrant digital public sphere while mandating some of the shortest content-removal deadlines in the world? Or does speed become a governing ideology—where consultation is bypassed, ambiguity is operationalised and compliance substitutes for deliberation?
The Modi government argues that in an age where information moves at digital velocity, State regulation must keep pace. Perhaps. But when regulation outpaces reflection, the clock does more than measure time. It redistributes authority.
Three hours is not merely a deadline. It is a declaration that in India’s digital sphere, the State has chosen to set the tempo—and expects everyone else to keep time. Indeed, an extension of its autocratic governance expeditions in other realms of life and society.
Citizens for Justice and Peace along with Sabrangindia has carried out a detailed clause by clause analysis of the new IT rules. Read the analysis here.
“Fayaz Altaf’s ‘Vendetta State’ is a sharp and thought-provoking piece that exposes uncomfortable truths about power and accountability. A timely story that demands serious reflection.”
Venkatesh Ramakrishnan presents a deeply researched and incisive analysis that cuts through the noise and lays bare the larger political and institutional implications. A compelling and courageous piece of journalism.”