Without transparency and independent oversight, takedown actions could infringe on democratic rights, particularly when they silence dissent.
There is a word that is increasingly becoming part of our digital vocabulary: takedown. Now imagine if the government had complete autonomy over what goes on your ‘social media’ pages. Most neutral observers would say that defeats the entire construct of social media which is supposed to be forums where people can socially express their opinion without the threat of censorship or govt controls… Would we then find ourselves being compared to our neighbour China a totalitarian regime, where social media is strictly monitored and dissent is wiped out as quickly as it is posted? The country has banned Facebook, Instagram, X and WhatsApp, as they ‘allegedly’ do not comply with local regulations. In a democracy like ours, being compared to China when it comes to expressing dissent would be a serious downfall. But the government is serious about social media and it is going after platforms. Meta, the parent company of Instagram and Facebook, recently got a dressing down from the Union government over what we know was the brief removal of a video message by Prime Minister Narendra Modi. But there are also sudden changes that users are talking about in discussion forums where they are being shown recommended posts from various creators. Instagram posts highlighting Gen Z dissent are getting removed with a simple message – We have taken this action based on a local legal requirement.
And as per data that The Indian Express has accessed, every 60 seconds, one blocking order is passed on average, making it 1.95 lakh orders in five months. Of the three platforms, Instagram, which the PM recently asked his ministers to join to boost their presence and reach among Gen Z, accounted for the largest share of these government directions, receiving nearly 1 lakh blocking orders during the five-month period, just over half of the total. Facebook, which has seen a slowdown in user growth worldwide, received around 80,000 orders, while YouTube received nearly 15,000. These figures count blocking orders, each of which can cover hundreds of individual pieces of content or accounts. A majority of these orders have been issued through the Home Ministry’s Sahyog portal, a platform developed to send blocking notices by various Central and state agencies.
According to the Home Ministry’s 2024-25 annual report, a little over 1.11 lakh pieces of ‘suspicious online content’ were blocked until March 2025 under Section 79(3)(b) of the Information Technology (IT) Act, including through the Sahyog portal, in the whole year. Meta, in order to comply with the government’s three-hour takedown mandate, has reportedly integrated its API with the Sahyog portal, where every flagged piece of content will automatically be taken down from its platform without any human review by the company. This removes Meta’s ability to contest the takedown directives. Now if this is true, an API executing takedowns without anybody at Meta reading and reviewing the concern makes it an unconditional compliance.
And when government is cutting down time of response, there needs to be someone watching the takedown machinery itself. There should be an independent review committee for the periodic examination of takedown directions, particularly when constitutionally protected speech is involved. He also calls for government agencies and intermediaries to publish periodic transparency reports containing aggregate information on takedown requests, along with a practical, time-bound mechanism through which affected users can seek reconsideration and judicial review.
When it comes to takedowns, we are told reasons like national security, public order, misinformation, harmful content or compliance with the law and some may be entirely legitimate. No democracy can argue that content inciting violence, facilitating terrorism or posing a genuine threat to public safety should remain online simply in the name of free speech. But there is a larger question we need to ask as governments acquire greater powers to demand faster removal of online content… When does a legitimate takedown become a violation of freedom of speech?
Because a takedown is not merely the removal of a piece of content, it is also the removal of someone’s ability to speak and everyone else’s ability to hear that speech. That matters enormously when the content being removed largely involves criticism of those in power. The government has a legitimate interest in protecting citizens and the country but it cannot also become the unquestioned judge of what citizens are allowed to say about the government itself.. that state is called curbing free speech something every ruler would want to but something that is against the very tenet of democracy.
Without an explanation, a takedown can become a black box. Content is released, a government direction goes out and the content disappears. That cannot be enough when the fundamental right at stake is freedom of expression.
Technology lawyer Deepak Singh argues that the answer lies in building safeguards into the takedown process itself. “First, there should be a clear takedown mechanism under the IT Act,” he says, including details of the designated officers responsible for the underlying offence being made available on the government’s website. There should also be clarity on ‘the extent to which adjudication of the content must be handled through judicial processes.’
That question of who gets to decide is crucial. The greater the power to remove speech from the public domain, the greater the need to ensure that such decisions are not made without accountability. Singh says authorities should also follow the principle of proportionality and use ‘the least restrictive means necessary to remove unlawful content’ rather than imposing broad restrictions that could affect lawful speech.
If one piece of content is unlawful, should an entire account be restricted? A takedown should be targeted at the unlawful act, not become a convenient way of silencing everything around it. There is also the question of giving people a chance to defend themselves. “Identified publishers and creators should ordinarily receive a written, reasoned notice and an opportunity to be heard,” Singh says. There can, of course, be exceptions. In an emergency involving national security or an imminent threat, waiting for a full hearing may not be possible. But even then such actions should be subjected to ‘prompt post-decisional review’, he adds. That is an important safeguard because extraordinary powers should not become permanent powers simply because they were exercised in the name of an emergency.
And there needs to be someone watching the takedown machinery itself. There should be an independent review committee for the periodic examination of takedown directions, particularly when constitutionally protected speech is involved. He also calls for government agencies and intermediaries to publish periodic transparency reports containing aggregate information on takedown requests, along with a practical, time-bound mechanism through which affected users can seek reconsideration and judicial review………
These may sound like procedural details but procedure is what protects a right from becoming a promise on paper. The concern becomes even greater when the machinery for takedowns is made faster and more efficient. Here speed may sound like efficiency but when it comes to speech, speed can also mean less time for scrutiny, less time for platforms to question a request and less time for the person affected to challenge the decision.
A mistaken order to remove a post can be corrected later. But the damage may already have been done. A story may have stopped circulating. A journalist’s reporting may have disappeared at the moment it mattered. A citizen’s criticism may have vanished during an important public debate & a fear of govt’s hard cold stare may have spread to a point where there would be no next expression of a free opinion something that all rulers want.
Also it can have a bigger chilling effect. If people know that content can be removed quickly and that challenging the decision is difficult, they may simply decide not to post in the first place. That is how censorship can work without requiring every dissenting voice to be formally banned as people begin censoring themselves.
This becomes particularly important in a country where social media has become a major arena for political conversation. India’s young population has embraced these platforms to discuss politics, education, jobs, governance and accountability. Gen Z has demonstrated that it can mobilise online and take its concerns into the real world and our political leaders naturally want to reach this audience.
Prime Minister Narendra Modi’s outreach to Gen Z on Instagram is a reflection of that changing political landscape. But there is an important democratic bargain here: if governments want to enter the spaces where young people communicate, they must also be willing to hear what young people have to say, even when that speech is critical. You cannot ask Gen Z to listen to the government on Instagram while making it increasingly difficult for Gen Z to question the government on Instagram. That is where the takedown debate becomes a free-speech debate. And this is why the comparison with China’s heavily controlled social-media environment should serve as a warning, not because India is China, but precisely because it is not.