Twitter may have blocked former US President Donald Trump for life on its platform but when it came to safeguarding the rights of those supporting farmer protests, it was forced to comply with a secretive, exclusionary process in Indian law that denies citizens the right to freedom of speech and expression
New Delhi: Microblogging website Twitter’s decision to toe the Indian government’s line and block accounts highlights the problems that social media platforms and citizens face in a process where accusations are secret, the complainant is unknown and decisions are taken behind closed doors.
On 31 January 2021, Twitter received an order passed under section 69A of the Information Technology Act (IT Act), 2000, to block 257 URLs and the hashtag “#ModiPlanningFarmerGenocide”. On 4 February 2021, it was served another order to block
Twitter argued that blocking accounts of activists, politicians and journalists would "violate their fundamental right to free expression under Indian law” before eventually complying with the government order.
If you were wondering how the same Twitter took action against former US President Donald Trump, banning him from the platform for life, remember two things: The Trump ban occurred under Twitter’s own policies without the involvement of a government and the ban was the reverse of what happened in India. Twitter took action against a former member of government, instead of in compliance with it.
In India, Twitter’s brief attempt to stand up for the right to free speech resulted in the government threatening action under section 69A(3) that carries a prison term of seven years and a fine.
The legal regime that governs the removal/blocking of information online is managed by officials of the Ministry of Electronics and Information Technology (MEITY). This article focuses on the procedural steps by which online speech is curbed, rather than the substantive law on which such curbs are justified.
Intermediaries In Law
Online platforms, including social media platforms such as Twitter, are “intermediaries”. They neither generate the information nor govern to whom it is delivered. They merely provide the platform on which it is distributed.
Intermediaries are given special protection against liability for third-party information they host, in exchange for compliance with certain conditions, including government directions to remove offending information under section 79(3)(b) of the IT Act.
Twitter itself is therefore not liable for any speech on its platform as long it takes down/blocks the offending speech once the government or a court orders it to do so.
Section 69A Blocking Procedure
The government’s power to direct the removal/blocking of information online stems from section 69A of the IT Act. The substantive grounds on which information can be blocked are if it is in the interests of:
1. Sovereignty and integrity of India
2. Defence of India
3. Security of the state
4. Friendly relations with foreign states
5. Public order
6. Preventing incitement for commission of a cognizable offence relating to any of the above categories.
These largely correlate with the restrictions in Article 19(2) of the Constitution. Even after accounting for certain overlap, section 69A grants the government power to block information on grounds that are narrower than those envisaged in the Constitution.
There are two steps to block information:
1. The government is satisfied that it is necessary or expedient to do so.
2. It records its reasons in writing.
The procedure was expanded upon in rules framed in 2009, called the Information Technology (Procedure and Safeguards for Blocking for Access of Information by Public) Rules, 2009. These rules provided for 3 routes to remove access/block information on the Internet:
1. An order passed by a competent court (Rule 10)
2. An order passed following procedure under Rules 7 and 8 of the 2009 Rules (called the “Regular Procedure” in this article)
3. An order passed under Rule 9 of the 2009 Rules (“Emergency Procedure”)
The second and third routes are adjudicated exclusively by government officials, and not by courts or judicial officers in a tribunal. A brief examination of each procedure, called the regular and emergency procedures for the purposes of this article, highlights the discrepancies that govern these restrictions on free speech, as opposed to the treatment it would have merited in open court.
The most salient of these differences is the lack of knowledge of the complainant/accuser or prior information of removal to the person who posted the offending information since all complaints and requests for removal of information are to be kept confidential under Rule 16 of the 2009 Rules.
Under the Regular Procedure, there are three levels of review and a right of personal hearing (for the intermediary, not content writer) before any order is passed for blocking information.
A complaint submitted moves from a nodal officer of a government agency/ministry to the central government’s designated officer. The latter chairs a committee of joint secretary-rank officers from the ministries of law and justice, home affairs, information and broadcasting and the Indian Computer Emergency Response Team.
The nodal officer and the committee examine whether the request falls under section 69A and whether any action is needed. The committee also makes all reasonable efforts to find the intermediary or person hosting the information in question (including foreign entities), who are given a right of hearing and reply before the committee.
The person who actually posted the information, as opposed to the intermediary/person hosting it, is not sought to be identified, nor is that person granted either a right of hearing or a right of reply (discussed in greater detail later in this article). The difference is important since people rarely own/host the platforms on which they express their opinions.
After this, the committee makes a recommendation in writing to the secretary of the department of information technology under MEITY, who must approve all requests for blocking information. The procedure outlined above finishes within 7 days from when the designated officer receives the request.
This procedure is activated on the fulfilment of two conditions:
1. It is a case of emergency nature
2. No delay is acceptable
Given that the restriction is being placed on a fundamental right of freedom of speech and expression and no right of personal hearing and reply exists for the intermediary or content writer/producer before information is blocked, stringent criteria has been put in place to qualify the nature of the emergency. The emergency must be of such immediate and urgent nature, that no delay in action is acceptable under the circumstances.
In this procedure, the nodal officer and the designated officer (there is no committee at this stage) separately assess and recommend whether the material is covered under section 69A and that it is “necessary and expedient” to block the information.
The secretary of the department of IT then assesses the material and, after recording reasons in writing, passes an interim order for blocking the information. The intermediary or content writer are not heard during this process.
Subsequent to this order, within 48 hours, the committee constituted under the regular procedure above considers the matter and gives its recommendation. The matter then reverts to the secretary of the department of information technology to pass a final order for approval/denial of the request to block information.
By the time the committee convenes, the emergent circumstances in which “no delay is acceptable” have already been dealt with by the interim order passed by the secretary of the department of IT. However, the person who posted (not hosted) the information or even the intermediary still do not have a right to be heard by the committee at this stage.
The committee makes its recommendations and the secretary of the department of IT passes the final order without either of them having heard the intermediary or content writer of the information in question.
Of the several blocking orders that Twitter received, by its own admission, at least two were emergency blocking orders, that Twitter temporarily complied with before restoring access to the accounts.
News reports suggest that MEITY was concerned about “incitement to genocide” being a “grave threat to public order”, perhaps believing the hashtag “#ModiPlanningFarmerGenocide” and therefore exercised its powers under section 69A to pass an emergency order for blocking information.
The grounds on which MEITY thought that, not only is this an emergency, but that “no delay is acceptable” are not in the public domain, and cannot be independently verified.
Ordered In Secret
Both procedures fail to consider the principal stakeholder in the discussion–the person who has actually posted the information. This is the person whose speech is infringed. The right to be heard and to present a defense is a core principle of natural justice, reiterated time and again without exception through a catena of Supreme Court decisions spanning half a century. Despite that, the person suffering the brunt of official action here has no right of hearing in either the emergency procedure or the regular procedure.
Given the fact that complaints/requests and “actions taken thereof” are kept “strictly confidential” under the 2009 Rules, a person exercising their right of free speech online can be:
1. Accused in secret
2. By person they do not know
3. Tried in absentia (i.e. without the person posting being present)
4. Without even a notice of proceedings
5. Denied a right to plead their case
6. Suffer removal of their speech
7. In an order passed by government officials (not judicial officers)
8. Denied access to an order, kept secret by law.
Even though Twitter mounted a short-lived defence of the right to freedom of speech and expression, that was an individual decision of one intermediary in a specific instance. The interests of the intermediary as a business concern operating in India and the interests of the person exercising freedom of speech online as a citizen can significantly diverge.
If the intermediary follows the bare letter of the law and quietly removes the offending post without notice to the person who wrote it, the intermediary will be protected by the liability shield under section 79 and the confidentiality rules that hide actions taken on complaints received. The person posting the content might never learn of its removal.
Beyond the infringement of an individual’s fundamental rights, these secret orders have notable societal effects. The first effect is that the public is unaware of the standards employed and the level of scrutiny applied to each complaint, before an executive order is passed directly infringing on the fundamental right of speech and expression.
The second effect is an inability to test the legality of executive action in censoring speech since the actual order is not published. The third effect is a corollary of the first and second–patterns of government malfeasance, if any, escape scrutiny since a large number of blocking orders cannot be studied together to identify the censorship regime deployed in its entirety on basis of what the government considers “necessary or expedient”.
The constitutional validity of the 2009 Rules was challenged in the Supreme Court in Shreya Singhal vs Union of India.
In the 2015 judgement that struck down section 66A of the IT Act, the 2009 Rules were held constitutionally valid as section 69A was considered to be a narrowly-drawn provision with several safeguards–the government must be satisfied that it is necessary to block the information, the necessity relates to the grounds set out in Article 19 (2) of the Constitution, and that the blocking order is in writing and can be challenged in a High Court.
Further, the Supreme Court held that the originator of the information is also to be heard pursuant to Rule 8, before an order for blocking information is passed. It is respectfully submitted that the originator of the information has no right of hearing under Rule 8, as it stands today.
The “person” mentioned in Rule 8 is the “person or intermediary who has hosted the information.” Often, and in every case involving social media platforms, the originator of the information i.e. the person who posts the information does not usually host it as the underlying cyber infrastructure is owned and maintained by the intermediary.
From available information, it appears that the government agrees with the latter interpretation. Even on the date when the Supreme Court gave the decision in Shreya Singhal, there was no publicly recorded instance of a pre-decision hearing being extended prior to passing of a section 69A order.
The Centre for Internet and Society reports that a blocking order passed for the satirical website Dowrycalculator.com in September 2019 did not afford a right of a pre-decision hearing despite the details of the creator of the website, Tanul Thakur, being widely known (separate from the issue that satire is not a ground for blocking information).
This order was passed more than four years after the decision in Shreya Singhal and it is evident that the government was in flagrant disregard of the Supreme Court’s ruling that the originator of the information should also be heard. There is no publicly available evidence that a right of hearing was extended to any of the owners of the Twitter accounts currently in question, including some with a public presence who were easy to identify and trace, such as the CEO of Prasar Bharti.
Even the person whose speech is restricted, Thakur in the dowry calculator website case, was denied a copy of the order by MEITY on the grounds of confidentiality under Rule 16, evidenced by the reply to the RTI application, made available by the Internet Freedom Foundation.
This is an impediment to challenging it in a writ petition before a High Court, a position of law that the Supreme Court relied upon while upholding the constitutionality of the 2009 Rules in Shreya Singhal, and reiterated in Anuradha Bhasin that an order restricting fundamental rights needs to be made public.
Under the 2009 Rules, a Review Committee meets every two months and examines all directions to block information to ensure compliance with section 69A. It also has the power to set aside directions, if it feels that they were exercised without cause. The minutes of the Review Committee meetings are not publicly available, so the extent and rigour of the review applied cannot be independently verified.
So Is India A Nation Of Cyber Terrorists?
Keeping aside the procedural weaknesses that afflict this process, a study of the government’s actions before and after these orders is also indicative of the seriousness with which it exercises its power to block information.
The grounds on which information is blocked are nearly exclusively related to national security–the sovereignty and integrity of India, the defence of India, the security of the state, friendly relations with other states and public order or for preventing incitement for commission of a cognizable offence relating to these.
It is important to note that preventing incitement to commission of an offence is also narrowly tailored by two criteria–one, the offence must be cognizable, and two, it must relate to one of the five substantive grounds, and not to any other offence.
It stands to reason that if speech is curbed on any of these grounds, there is a much larger context beyond the immediate offending post in question.
One post on Twitter alone will not pose a credible threat to the sovereignty or the integrity of India, or the defence of India or the security of the state or friendly relations with other states or public order.
There will be actions online and off-line that precede, accompany and succeed it. The presence of a credible threat is safe to assume predicated on the understanding that senior officers of the government do not trouble themselves with every trivial post that appears in the far corners of the Internet. The law provides for what these actions can look like.
The commission of a cognizable offence relating to the above is covered under the provisions relating to cyber terrorism in section 66F, with the exception of public order that was covered in the now repealed section 66A.
Section 66F lays down a stringent and detailed procedure for determining what constitutes cyber terrorism, and in relation to the four grounds named above, requires the following actions to have occurred:
1. A knowing or intentional penetration or access of a computer resource;
2. The absence of authorisation or exceeding the authorisation given;
3. Obtaining information or data
a. restricted for reasons of security of state or foreign relations, or;
b. knowing that the data can harm the sovereignty and integrity of India, the security of the state, friendly relations with other states, public order and certain other grounds.
The other criteria relating to some of the grounds, including unity, integrity, security and sovereignty of India alternatively have separate criteria for deciding whether a case of cyber terrorism has occurred:
1. Intention to threaten the unity, integrity, security and sovereignty of India
2. Denial of access to a computer resource, unauthorised penetration of a computer, or introduction of a computer virus/contaminant
3. And this conduct must be of a severity likely to cause death or injury to a person or destruction of property, or damage or disrupt essential community supplies or affect critical information infrastructure.
The offence of cyber terrorism is punishable with imprisonment for life. Several hostile actions need to have occurred, beyond a simple post on Facebook or Twitter–obtaining restricted information, denying access to an important computer resource with significant catastrophic results, spreading computer viruses–for cyber terrorism to take place.
This is the larger context which makes an online post indicative of future harm–the credible threat to which the government responds. If none of these have occurred or are occurring, that post will fall into the category of one of thousands of trivial posts with which senior government officers do not worry themselves. This context supplies the basis on which ostensible government censorship of speech is justified.
This is the larger context, and not the post, which will justify the invocation of an emergency procedure for blocking online information under section 69A for which, as discussed, two criteria are necessary:
1. The case is of emergency nature
2. No delay is acceptable
It then stands to reason that this larger context is investigated by the government and prosecuted accordingly. For instance, 650 posts on Facebook and 174 on Instagram were removed between January and July 2020, pursuant to section 69A orders, since the content was against “security of the state and public order”. Cumulatively, 2014 posts on Facebook and 77 posts on Instagram were removed in 2019, to comply with section 69A orders.
Comparable figures for Google across its products for reasons relating to “national security” are 156 in the first half of 2020 and 141 for all of 2019. Twitter’s data was not as disaggregated but the figures are still indicative of government activity.
Twitter received 2,768 legal requests for removal from the government in the first half of 2020 which was twice the number of requests that it received for all of 2019: 1,271.These numbers are of government requests only, and do not include requests by private individuals or court orders directing removal of information.
Twitter’s graph of receipt of legal requests to remove information is indicative of increased interest in this avenue over the years, with a nearly fourfold spike in the first half of 2020 compared to the second half of 2019.
Separately, and only on the blocking of URLs, government responses to parliamentary questions submitted in March 2020 show an increase from 1,385 in 2017 to 2,799 (2018) to 3,635 (2019). For 2020, that number was 9,849.
As we previously noted, the security of the state is not ensured by removing a Facebook post alone. It is likely that investigations and prosecutions were launched connected with these activities to protect national security.
For these 3,748 posts across three platforms alone (Google, Facebook/Instagram and Twitter), not counting other intermediaries who presumably also received orders to take down information, National Crime Records Bureau (NCRB) data shows that just 12 investigations for cyber terrorism were launched all over India in 2019, a sharp decrease from 23 in 2018. Just 199 crimes had terrorist activity applied as a motive and about 28 sought to disrupt public service.
Of the 35 investigations into cyber terrorism pending at the end of 2019, six went nowhere—two were found false, one ended because of a mistake of law or fact or civil dispute, and three were dropped for lack of evidence despite being found prima facie true.
Chargesheets were filed in only three and those trials are pending (in addition to 23 trials from previous years). Spliced another way, after blocking more than 3,500 online posts across just three platforms from which data is available, only 17 arrests were made of whom 11 faced trial, and none have been acquitted/convicted (as of early-mid 2020).
The other head under which these crimes are prosecuted is sedition. NCRB data show a cumulative 229 investigations in 2019 (inclusive of preceding years), of which 29 ended at the investigation stage itself, mostly for lack of evidence: 40 chargesheets were filed, and investigation remained pending in the rest.
In the 30 decisions that courts gave, there was one conviction and 29 acquittals. Efforts made by the government to investigate and prosecute offences impacting the sovereignty and integrity of India and other important considerations of national security resulted in an average conviction rate of 3.3%. Investigation is still pending in 159 cases.
The lack of effective follow-up government action, after blocking speech, to actually investigate and prosecute what it considers credible threats to the sovereignty and integrity of India, defence of India, security of the state, friendly relations with foreign states and public order leads to one of two irresistible conclusions.
One, that no credible threat actually exists/existed and therefore there was no cause to restrict the right to freedom of speech and use secret orders to block speech online.
The other possible conclusion is that the government has no interest in combating credible threats to the sovereignty and integrity of India and other important matters of national security, despite at one time believing that “no delay is acceptable” and invoking plenary emergency powers to block a tweet.
The principles of natural justice are discarded in their entirety by any single operation of the government procedure to block information online. Citizens are denied their fundamental right to freedom of speech and expression online by a process designed to be started, conducted and completed in secret. Currently, it is impossible to know if it was ever justified because “strict confidentiality shall be maintained”.
(Divyam Nandrajog is a Delhi-based lawyer.)