For the better part of a decade, the evolution of India’s technology regulations has been a noisy, public affair. Drafts were released, comments were invited, and stakeholder submissions, for all their technical jargon, were posted online for public scrutiny. It was a process that, while often contentious, was fundamentally transparent. That era may be over. In March 2026, the Ministry of Electronics and Information Technology (MeitY) released draft amendments to the Information Technology Rules, 2021, and with them, a quiet but profound change in procedure. All stakeholder feedback, the ministry declared, would be held in a “fiduciary capacity” and would “not be disclosed to anyone at any stage.”

The doors to the policy-making chamber have been closed. For hundreds of startups, tech platforms, and AI companies, this shift from public debate to private deliberation is more than just a procedural tweak. It is a fundamental alteration of the regulatory landscape. Without access to the arguments being made by industry peers, civil society, and Big Tech, founders are left flying blind, trying to anticipate rules that are being forged in secret. The key question is no longer just what the new rules will be, but how they are being made, and whose voices are being heard behind the curtain.

This move creates significant uncertainty for an ecosystem that thrives on predictability. The contents of these secret submissions, which we can piece together from conversations across the industry, reveal a deep-seated anxiety about the future of the internet in India, particularly concerning the foundational principles of intermediary liability, the burgeoning field of artificial intelligence, and the very viability of small, innovative tech companies.

What’s at Stake: The Unseen Amendments

While the final text remains unknown, the focus of the March 2026 draft amendments continues a regulatory trajectory that MeitY has been on for years: tightening the government’s control over online content and platform accountability. Based on the ministry’s recent focus areas, particularly around AI and misinformation, the proposed changes likely revolve around a few critical, and controversial, themes.

Expanding Liability for AI-Generated Content

Following global and domestic panic over sophisticated deepfakes and AI-driven misinformation campaigns, it is almost certain the amendments aim to place a heavier compliance burden on platforms and creators of generative AI tools. The government’s previous advisories, which urged platforms to ensure users do not post unlawful content, are likely being codified into hard law. This could translate into a mandate for platforms to proactively detect and filter AI-generated media that is deemed misleading or harmful, moving far beyond the current notice-and-takedown regime.

For startups building AI models or offering AI-as-a-service APIs, the implications are profound. The amendments could seek to hold them partially liable for the outputs created by their users, a move that would dramatically increase their legal exposure and compliance costs.

Strengthening the Grievance Appellate Committee (GAC)

The GAC, established to hear user appeals against platform content moderation decisions, is another likely area of focus. The amendments could expand the GAC’s powers, potentially giving it the authority to levy fines or issue binding orders that go beyond individual content pieces to dictate broader platform policies. While intended to empower users, many platforms view the GAC as a state-run body with the power to overrule their internal safety mechanisms, creating a conflict with their global content policies and community standards.

Shorter Takedown Timelines and Broader Definitions

The existing IT Rules already mandate specific timelines for acknowledging user grievances and acting on court orders. The new amendments are expected to shorten these timelines further, particularly for content flagged by the government or its designated agencies. Furthermore, there is a persistent push to broaden the definitions of what constitutes “patently false,” “misleading,” or “harmful” content, giving government agencies more latitude to order takedowns and platforms less room to exercise judgment.

The Industry’s Muffled Pushback

With the formal consultation process now a black box, the industry’s response is happening through private letters and closed-door meetings. However, the core arguments being made against the government’s direction are consistent across the board, from the largest social media giants to the smallest content-sharing startups.

The Slow Death of Safe Harbour

The most significant concern is the erosion of “safe harbour” protections under Section 79 of the IT Act. This legal principle protects intermediaries, such as social media platforms, e-commerce sites, and cloud providers, from being held liable for the content posted by their users, provided they follow certain due diligence requirements. It is the legal bedrock that allows the user-generated internet to exist.

Any rule that requires a platform to proactively monitor or filter content before it is published effectively negates this protection. It shifts the platform’s role from a neutral intermediary to a proactive publisher, with all the attendant legal risks.

For large companies like Google or Meta, this is a massive, expensive compliance challenge. For an Indian startup building a new community platform or a SaaS tool that allows user collaboration, it is an existential threat. They lack the capital to invest in the sophisticated AI filtering tools and the armies of human moderators required for such a task. The result, industry bodies argue in their submissions, would be a chilling effect on innovation and a consolidation of power in the hands of a few Big Tech players who can afford the cost of proactive censorship.

Crippling AI Innovation Before It Starts

The push to regulate generative AI at the platform level is seen by many in the AI startup community as a critical misstep. The global consensus, reflected in frameworks like the EU AI Act, is moving towards a risk-based approach. This means regulating the specific application of AI (e.g., in critical infrastructure or law enforcement) rather than the underlying technology itself.

By contrast, India’s potential approach of holding AI model creators liable for downstream uses would be an outlier. It would make it incredibly risky for an Indian startup to build and release a foundational AI model. Investors would be wary of funding companies with such an open-ended and unpredictable legal liability. The argument being made to MeitY is that this would force Indian innovation offshore and make the country a net importer of AI technology, completely at odds with the government’s “Make in India” ambitions.

The Problem of Opaque Policymaking

Beyond the substance of the rules, the process itself has become a major point of contention. The decision to keep submissions confidential is being framed by legal experts and civil society groups as a violation of the principles of natural justice and the Pre-Legislative Consultation Policy of 2014. Public consultation is not just about gathering feedback; it is about allowing stakeholders to see and respond to each other’s arguments, leading to more robust and well-rounded policy.

When submissions are secret, it is impossible to know if the government is genuinely considering a wide range of views or simply cherry-picking arguments that support a predetermined outcome. This opacity undermines trust in the regulatory process and makes it difficult for businesses to plan for the future with any degree of confidence.

The Founder’s Playbook: Navigating the Uncertainty

With the final rules still pending and the debate happening out of public view, what should a founder or a compliance head do? Waiting for the final notification is not a viable strategy. Proactive steps must be taken now.

  • Audit Your Safe Harbour Compliance: Go back to the basics of Section 79. Ensure your terms of service, user policies, and grievance redressal mechanisms are not just compliant with the 2021 rules, but are robust enough to handle a stricter regime. Document every step of your content moderation process meticulously. This documentation will be your best defense if your safe harbour status is challenged.
  • Scenario Plan for AI Liability: If you are an AI-first company, you must game out the worst-case scenarios. What if you are held liable for a deepfake created using your API? Work with your legal team to draft stronger indemnity clauses in your user agreements. Invest in developing technical guardrails and content filters for your own tools, even if they are not yet legally mandated. Demonstrating proactive effort can be a powerful mitigating factor.
  • Leverage Industry Associations: In an environment of opaque consultation, collective action becomes paramount. Your voice is louder when it is part of a chorus. Actively engage with industry bodies like NASSCOM, IAMAI, or other tech policy groups. They have the access and resources to engage with MeitY directly, and their consolidated feedback, representing hundreds of companies, carries more weight than a single, secret submission.

A Regulatory Fog Descends

The government’s intent is clear: to create a safer, more accountable digital ecosystem. It is a laudable goal that most in the industry share. However, the path being taken, one of increasing state control and decreasing transparency, risks achieving the opposite. By creating a climate of regulatory uncertainty and fear, it may stifle the very innovation and competition that makes the digital ecosystem dynamic.

The final version of the 2026 IT Rules amendments, when they are eventually notified, will be a crucial signal. It will tell us whether the government has listened to the hushed, urgent warnings from its own startup ecosystem. More importantly, the process has set a worrying precedent. This shift towards opaque consultation could prove more damaging to the “ease of doing business” and India’s ambition to be a global tech leader than any single burdensome regulation. For now, the ecosystem waits in a regulatory fog, hoping for a clear sky that may not appear.