India Just Gave Social Media Companies Three Hours to Delete Content – Or Else

Imagine you're running a platform with over a billion users. You get a government notice at 2 AM saying some content violates the law. You have three hours to find it, review it, verify the order's legitimacy, remove it, and document everything. Miss that deadline and you lose the legal shield that keeps you from being held liable for everything your users post.

Well, that's no hypothetical situation any longer. That's India's new reality starting February 20.

A couple of days ago, the Ministry of Electronics and Information Technology quietly dropped gazette notification G.S.R. 120(E), slashing content takedown timelines from 36 hours to three. Meta, X, and YouTube were given exactly 10 days to figure out how to comply. In other words, censorship in India just hopped on some serious steroids.

When "Fast" Becomes "Instant Censorship"

To begin with, the notification updates the IT Rules, 2021. Platforms that previously had 36 hours to act on government takedown orders now get three. For non-consensual intimate imagery and deepfakes, it drops to two hours.

India already issues thousands of takedown orders annually. Meta restricted more than 28,000 pieces of content in India during the first half of 2025 alone. Now every decision needs to happen 12 times faster.

The government didn't explain why three hours. They circulated a draft for consultation in October 2025, but the final notification arrived without public discussion. Akash Karmakar, a partner at law firm Panag & Babu, called it "practically impossible" to implement with any meaningful review.

The Automation Trap Nobody's Talking About

Three hours might sound reasonable until you realize what it means in practice. Platforms serve over a billion Indian users across multiple time zones and languages. To meet a three-hour deadline reliably, you need automation. Full, aggressive, automated enforcement with zero delays.

The Internet Freedom Foundation warns that this timeline "eliminates any meaningful human review," forcing platforms toward automated over-removal. Technology analyst Prasanto K. Roy called it "perhaps the most extreme takedown regime in any democracy."

Platforms can't assess whether requests are legally appropriate. They can't evaluate context. They can't distinguish between genuine violations and contested speech. When you have three hours to decide between massive fines and removing content, you remove the content every time.

And once India establishes this as standard, other governments will demand the same instant compliance globally. We’ve already seen this pattern, it’s really nothing new.

Ten Days To Build A Censorship Machine

February 20 isn't far away. Platforms were given exactly 10 days to build or reconfigure systems that can process government orders in three hours. And, of course, given the situation, they won't be building thoughtful review systems but automated kill switches.

Expect a massive spike in content removals starting February 20. Not because users suddenly posted more illegal material, but because platforms will err on deletion. Anything remotely questionable disappears because for all these companies, it’s always better to be censored than sued.

The casualties won't just be harmful content. They'll include political criticism, contested facts, and anything algorithms flag as risky. When speed becomes the only performance metric, accuracy dies first. This is what happens when compliance timelines shrink to the point where only machines can meet them.