Content moderation is a wicked problem - that calls on us to balance the fundamental right to speech and expression with the need to prevent the spread of harmful content. This is a delicate balance that we need to strike with careful attention to constitutional principles.

This is a link-enhanced version of an article that first appeared in the Mint. You can read the original here. For the full archive of all the Ex Machina articles, please visit the website.

In a petition before the Karnataka high court, X (formerly Twitter) has challenged the legality of India’s Sahyog Portal—a central digital repository that tells social media companies which sites they must block access to. This, the company argues, is an attempt to side-step procedures already approved by the Supreme Court under Section 69A of the Information Technology Act, 2000 (IT Act).

The Content Moderation Debate

Few would disagree that some online content (child pornography, trafficking, etc) is so utterly unacceptable that law enforcement must actively track it down and immediately remove all traces of its existence.  At the same time, there is other content that some may want removed for ideological, personal or other less explicable reasons. This content should not be taken down unless the reasons for doing so align with one of the restrictions on freedom of speech—the sovereignty and integrity of India, security of the State, public order, etc —that have been set out under Article 19(2) of the Indian Constitution. This is the central tension in India’s content-moderation debate. While we must empower law enforcement agencies to take down harmful content, we must ensure they do not misuse this power and stifle our fundamental right to speech and expression.

It would be nice if the government could be relied on to take a constitutionally aligned approach to content moderation. Unfortunately, this has not been our experience. In their haste to prevent harmful content from spreading, law enforcement agencies show uncommon haste in riding roughshod over fundamental rights, often adopting expansive interpretations of the restrictions under Article 19(2) so that they can take down anything even remotely uncomfortable to them.

It was in order to legitimize this sort of broad executive action that Section 66A was introduced into the IT Act, with a view to prosecuting those who send ‘offensive messages’ online. This was challenged before the Supreme Court in the Shreya Singhal vs Union of India case and struck down as violative of the right to freedom of speech. Shreya Singhal also questioned the constitutional validity of Section 69A, which gave the government the authority to block access to online information, alleging that the terms used in that section were so broad that the government could restrict speech on the flimsiest of grounds.

On this, the Supreme Court disagreed, arguing that Section 69A mirrored the restrictions set out in Article 19(2) of the Constitution and, as such, merely reiterated existing constitutional exceptions. Since blocking orders had to be in writing, they could be challenged in court, and there was a detailed procedure that all blocking requests had to adhere to. The court also drew comfort from the fact that the related Rules had constituted a review committee that met every two months to ensure that the blocking orders that had been issued were in accordance with the requirements of the IT Act.

Intermediary Liability

The latest constitutional challenge is on account of the fact that the Government of India has taken to relying on Section 79 to block online content.

The original purpose of this section was to protect online intermediaries like X from being held liable for content that users post on their platforms. This ‘safe harbour’ exemption protects companies that host user-generated content from being sued for material over which they have no control but is only available to them if they expeditiously take down all content that the government notifies them for removal. It is this condition that the government has begun to exploit in order to enforce content takedowns and for which the Sahyog portal was created.

Online platforms receive takedown notices from all sorts of government officials—so much so that it is almost impossible for them to figure out whether a given request is genuine. I have personally seen requests that purport to come from police inspectors in remote corners of the country but have been sent from generic email accounts that cannot be traced to an identifiable official or department. The Sahyog Portal is an attempt to address these ambiguities. A central digital repository that has a definitive list of all the sites notified by the government for blocking reduces the scope for confusion over authenticity.

Platform for Transparency

This is a significant benefit. If intermediaries only have to block sites that are listed on the portal, they can ignore the hundreds of random requests they currently receive. Also, with all sites that have been notified as needing to be blocked listed in a single location, it is much easier for free-speech activists to assess the constitutional validity of the blocking orders made by the government in each such instance. A portal for transparency around take-downs can only be a good thing.

The trouble is that neither the Sahyog Portal nor Section 79 has any of the safeguards that were built into Section 69A, which allowed the Supreme Court to uphold its constitutional validity. Unless the websites that have to be blocked are notified on the Sahyog Portal according to a procedure that complies with Article 19(2) restrictions, it will likely be held unconstitutional. What’s more, all such notices should also be subject to review by the same review committee that currently supervises takedowns under Section 69A of the IT Act.

The Supreme Court has already told us how content should be taken down. The government just needs to comply.