Centre for Internet & Society

Most software code has dependencies. Simple and reproducible methods exist for mapping and understanding the impact of these dependencies. Legal code also has dependencies --across court orders and within a single court order. And since court orders are not produced using a structured mark-up language, experts are required to understand the precedential value of a court order.

The article was published in the Economic and Political Weekly Vol-L No.15.  Vidushi Marda, programme officer at the Centre for Internet and Society, was responsible for all the research that went into this article. PDF version here.

As a non–lawyer and engineer, I cannot authoritatively comment on the Supreme Court’s order in Shreya Singhal vs Union of India (2015) on sections of the Information Technology Act of 2000, so I have tried to summarise a variety of views of experts in this article. The Shreya Singhal order is said to be unprecedented at least for the last four decades and also precedent setting as its lucidity, some believe, will cause a ripple effect in opposition to a restrictive understanding of freedom of speech and expression, and an expansiveness around reasonable restrictions. Let us examine each of the three sections that the bench dealt with.

The Section in Question

Section 66A of the IT Act was introduced in a hastily-passed amendment. Unfortunately, the language used in this section was a pastiche of outdated foreign laws such as the UK Communications Act of 2003, Malicious Communications Act of 1988 and the US Telecommunications Act, 1996.1 Since the amendment, this section has been misused to make public examples out of innocent, yet uncomfortable speech, in order to socially engineer all Indian netizens into self-censorship.2

Summary: The Court struck down Section 66A of the IT Act in its entirety holding that it was not saved by Article 19(2) of the Constitution on account of the expressions used in the section, such as "annoying," "grossly offensive," "menacing,", "causing annoyance." The Court justified this by going through the reasonable restrictions that it considered relevant to the arguments and testing them against S66A. Apart from not falling within any of the categories for which speech may be restricted, S66A was struck down on the grounds of vagueness, over-breadth and chilling effect. The Court considered whether some parts of the section could be saved, and then concluded that no part of S66A was severable and declared the entire section unconstitutional. When it comes to regulating speech in the interest of public order, the Court distinguished between discussion, advocacy and incitement. It considered the first two to fall under the freedom of speech and expression granted under Article 19(1)(a), and held that it was only incitement that attracted Article 19(2).

Between Speech and Harm

Gautam Bhatia, a constitutional law expert, has an optimistic reading of the judgment that will have value for precipitating the ripple effect. According to him, there were two incompatible strands of jurisprudence which have been harmonised by collapsing tendency into imminence.3 The first strand, exemplified by Ramjilal Modi vs State of UP4 and Kedar Nath Singh vs State of Bihar,5 imported an older and weaker American standard, that is, the tendency test, between the speech and public order consequences. The second strand exemplified byRam Manohar Lohia vs State of UP,6 S Rangarajan vs P Jagjivan Ram,7 andArup Bhuyan vs Union of India,8 all require greater proximity between the speech and the disorder anticipated. In Shreya Singhal, the Supreme Court held that at the stage of incitement, the reasonable restrictions will step in to curb speech that has a tendency to cause disorder. Other experts are of the opinion that Justice Nariman was doing no such thing, and was only sequentially applying all the tests for free speech that have been developed within both these strands of precedent. In legal activist Lawrence Liang's analysis, "Ramjilal Modi was decided by a seven judge bench and Kedarnath by a constitutional bench. As is often the case in India, when subsequent benches of a lower strength want to distinguish themselves from older precedent but are unable to overrule them, they overcome this constraint through a doctrinal development by stealth. This is achieved by creative interpretations that chip away at archaic doctrinal standards without explicitly discarding them."9

Compatibility with US Jurisprudence

United States (US) jurisprudence has been imported by the Indian Supreme Court in an inconsistent manner. Some judgments hold that the American first amendment harbours no exception and hence is incompatible with Indian jurisprudence, while other judgments have used American precedent when convenient. Indian courts have on occasion imported an additional restriction beyond the eight available in 19(2)-the ground of public interest, best exemplified by the cases of K A Abbas10 and Ranjit Udeshi.11 The bench in its judgment-which has been characterised by Pranesh Prakash as a masterclass in free speech jurisprudence12-clarifies that while the American first amendment jurisprudence is applicable in India, the only area where a difference is made is in the "sub serving of general public interest" made under the US law. This eloquent judgment will hopefully instruct judges in the future on how they should import precedent from American free speech jurisprudence.

Article 14 Challenge

The Article 14 challenge brought forward by the petitioners contended that Section 66A violated their fundamental right to equality because it differentiated between offline and online speech in terms of the length of maximum sentence, and was hence unconstitutional. The Court held that an intelligible differentia, indeed, did exist. It found so on two grounds. First, the internet offered people a medium through which they can express views at negligible or no cost. Second, the Court likened the rate of dissemination of information on the internet to the speed of lightning and could potentially reach millions of people all over the world. Before Shreya Singhal, the Supreme Court had already accepted medium-specific regulation. For example in K A Abbas, the Court made a distinction between films and other media, stating that the impact of films on an average illiterate Indian viewer was more profound than other forms of communication. The pessimistic reading of Shreya Singhal is that Parliament can enact medium-specific law as long as there is an intelligible differentia which could even be a technical difference-speed of transmission. However, the optimistic interpretation is that medium-specific law can only be enacted if there are medium-specific harms, e g, phishing, which has no offline equivalent. If the executive adopts the pessimistic reading, then draconian sections like 66A will find their way back into the IT Act. Instead, if they choose the optimistic reading, they will introduce bills that fill the regulatory vacuum that has been created by the striking down of S66A, that is, spam and cyberbullying.

Section 79

Section 79 was partially read down. This section, again introduced during the 2008 amendment, was supposed to give legal immunity to intermediaries for third party content by giving a quick redressal for those affected by providing a mechanism for takedown notices in the Intermediaries Guidelines Rules notified in April 2011. But the section and rules had enabled unchecked invisible censorship13 in India and has had a demonstrated chilling effect on speech14 because of the following reasons:

One, there are additional unconstitutional restrictions on speech and expression. Rule 3(2) required a standard "rules and regulation, terms and condition or user agreement" that would have to be incorporated by all intermediaries. Under these rules, users are prohibited from hosting, displaying, uploading, modifying, publishing, transmitting, updating or sharing any information that falls into different content categories, a majority of which are restrictions on speech which are completely out of the scope of Article 19(2). For example, there is an overly broad category which contains information that harms minors in any way. Information that "belongs to another person and to which the user does not have any right to" could be personal information or could be intellectual property. A much better intermediary liability provision was introduced into the Copyright Act with the 2013 amendment. Under the Copyright Act, content could be reinstated if the takedown notice was not followed up with a court order within 21 days.15 A counter-proposal drafted by the Centre for Internet and Society for "Intermediary Due Diligence and Information Removal," has a further requirement for reinstatement that is not seen in the Copyright Act.16

Two, a state-mandated private censorship regime is created. You could ban speech online without approaching the court or the government. Risk-aversive private intermediaries who do not have the legal resources to subjectively determine the legitimacy of a legal claim err on the side of caution and takedown content.

Three, the principles of natural justice are not observed by the rules of the new censorship regime. The creator of information is not required to be notified nor given a chance to be heard by the intermediary. There is no requirement for the intermediary to give a reasoned decision.

Four, different classes of intermediaries are all treated alike. Since the internet is not an uniform assemblage of homogeneous components, but rather a complex ecosystem of diverse entities, the different classes of intermediaries perform different functions and therefore contribute differently to the causal chain of harm to the affected person. If upstream intermediaries like registrars for domain names are treated exactly like a web-hosting service or social media service then there will be over-blocking of content.

Five, there are no safeguards to prevent abuse of takedown notices. Frivolous complaints could be used to suppress legitimate expressions without any fear of repercussions and given that it is not possible to expedite reinstatement of content, the harm to the creator of information may be irreversible if the information is perishable. Transparency requirements with sufficient amounts of detail are also necessary given that a human right was being circumscribed. There is no procedure to have the removed information reinstated by filing a counter notice or by appealing to a higher authority.

The judgment has solved half the problem by only making intermediaries lose immunity if they ignore government orders or court orders. Private takedown notices sent directly to the intermediary without accompanying government orders or courts order no longer have basis in law. The bench made note of the Additional Solicitor General's argument that user agreement requirements as in Rule 3(2) were common practice across the globe and then went ahead to read down Rule 3(4) from the perspective of private takedown notices. One way of reading this would be to say that the requirement for standardised "rules and regulation, terms and condition or user agreement" remains. The other more consistent way of reading this part of the order in conjunction with the striking down of 66A would be to say those parts of the user agreement that are in violation of Article 19(2) have also been read down.

This would have also been an excellent opportunity to raise the transparency requirements both for the State and for intermediaries: for (i) the person whose speech is being censored, (ii) the persons interested in consuming that speech, and (iii) the general public. It is completely unclear whether transparency in the case of India has reduced the state appetite for censorship. Transparency reports from Facebook, Google and Twitter claim that takedown notices from the Indian government are on the rise.17 However, on the other hand, the Department of Electronics and Information Technology (DEITY) claims that government statistics for takedowns do not match the numbers in these transparency reports.18 The best way to address this uncertainty would be to require each takedown notice and court order to be made available by the State, intermediary and also third-party monitors of free speech like the Chilling Effects Project.

Section 69A

The Court upheld S69A which deals with website blocking, and found that it was a narrowly-drawn provision with adequate safeguards, and, hence, not constitutionally infirm. In reality, unfortunately, website blocking usually by internet service providers (ISPs) is an opaque process in India. Blocking under S69A has been growing steadily over the years. In its latest response to an RTI (right to information)19 query from the Software Freedom Law Centre, DEITY said that 708 URLs were blocked in 2012, 1,349 URLs in 2013, and 2,341 URLs in 2014. On 30 December 2014 alone, the centre blocked 32 websites to curb Islamic State of Iraq and Syria propaganda, among which were "pastebin" websites, code repository (Github) and generic video hosting sites (Vimeo and Daily Motion).20 Analysis of leaked block lists and lists received as responses to RTI requests have revealed that the block orders are full of errors (some items do not exist, some items are not technically valid web addresses), in some cases counter speech which hopes to reverse the harm of illegal speech has also been included, web pages from mainstream media houses have also been blocked and some URLs are base URLs which would result in thousands of pages getting blocked when only a few pages might contain allegedly illegal content.21

Pre-decisional Hearing

The central problem with the law as it stands today is that it allows for the originator of information to be isolated from the process of censorship. The Website Blocking Rules provide that all "reasonable efforts" must be made to identify the originator or the intermediary who hosted the content. However, Gautam Bhatia offers an optimistic reading of the judgment, he claims that the Court has read into this "or" and made it an "and"-thus requiring that the originator must also be notified of blocks when he or she can be identified.22


Usually, the reasons for blocking a website are unknown both to the originator of material as well as those trying to access the blocked URL. The general public also get no information about the nature and scale of censorship unlike offline censorship where the court orders banning books and movies are usually part of public discourse. In spite of the Court choosing to leave Section 69A intact, it stressed the importance of a written order for blocking, so that a writ may be filed before a high court under Article 226 of the Constitution. While citing this as an existing safeguard, the Court seems to have been under the impression that either the intermediary or the originator is normally informed, but according to Apar Gupta, a lawyer for the People's Union for Civil Liberties, "While the rules indicate that a hearing is given to the originator of the content, this safeguard is not evidenced in practice. Not even a single instance exists on record for such a hearing."23 Even worse, block orders have been unevenly implemented by ISPs with variations across telecom circles, connectivity technologies, making it impossible for anyone to independently monitor and reach a conclusion whether an internet resource is inaccessible as a result of a S69A block order or due to a network anomaly.

Rule 16 under S69A requires confidentiality with respect to blocking requests and complaints, and actions taken in that regard. The Court notes that this was argued to be unconstitutional, but does not state their opinion on this question. Gautam Bhatia holds the opinion that this, by implication, requires that requests cannot be confidential. Chinmayi Arun, from the Centre for Communication Governance at National Law University Delhi, one of the academics supporting the petitioners, holds the opinion that it is optimism carried too far to claim that the Court noted the challenge to Rule 16 but just forgot about it in a lack of attention to detail that is belied by the rest of the judgment.

Free speech researchers and advocates have thus far used the RTI Act to understand the censorship under S69A. The Centre for Internet and Society has filed a number of RTI queries about websites blocked under S69A and has never been denied information on grounds of Rule 16.24 However, there has been an uneven treatment of RTI queries by DEITY in this respect, with the Software Freedom Law Centre25 being denied blocking orders on the basis of Rule 16. The Court could have protected free speech and expression by reading down Rule 16 except for a really narrow set of exceptions wherein only aggregate information would be made available to affected parties and members of the public.


In Shreya Singhal, the Court gave us great news: S66A has been struck down; good news: S79(3) and its rules have been read down; and bad news: S69A has been upheld. When it comes to each section, the impact of this judgment can either be read optimistically or pessimistically, and therefore we must wait for constitutional experts to weigh in on the ripple effect that this order will produce in other areas of free speech jurisprudence in India. But even as free speech activists celebrate Shreya Singhal, some are bemoaning the judgment as throwing the baby away with the bathwater, and wish to reintroduce another variant of S66A. Thus, we must remain vigilant.


1 G S Mudur (2012): "66A 'Cut and Paste Job,'" The Telegraph, 3 December, visited on 3 April, 2015, http://www.telegraphindia.com/1121 203/jsp/frontpage/story_16268138.jsp

2 Sunil Abraham (2012): "The Five Monkeys and Ice Cold Water," Centre for Internet and Society, 26 September, visited on 3 April 2015, http://cis-india.org/internet-governance/www-deccan-chronicle-sep-16-201...

3 Gautam Bhatia (2015): "The Striking Down of 66A: How Free Speech Jurisprudence in India Found Its Soul Again," Indian Constitutional Law and Philosophy, 26 March, visited on 4 April 2015, https://indconlawphil.wordpress.com/2015/03/26/the-striking-down-of-sect...

4 Ramjilal Modi vs State of UP, 1957, SCR 860.

5 Kedar Nath Singh vs State of Bihar, 1962, AIR 955.

6 Ram Manohar Lohia vs State of UP, AIR, 1968 All 100.

7 S Rangarajan vs P Jagjivan Ram, 1989, SCC(2), 574.

8 Arup Bhuyan vs Union of India, (2011), 3 SCC 377.

9 Lawrence Liang, Alternative Law Forum, personal communication to author, 6 April 2015.

10 K A Abbas vs Union of India, 1971 SCR (2), 446.

11 Ranjit Udeshi vs State of Maharashtra,1965 SCR (1) 65.

12 Pranesh Prakash (2015): "Three Reasons Why 66A Verdict Is Momentous"/ Times of India/(29 March). Visited on 6 April 2015, http://timesofindia.indiatimes.com/home/sunday-times/all-that-matters/Th...

13 Pranesh Prakash (2011): "Invisble Censorship: How the Government Censors Without Being Seen," The Centre for Internet and Society, 14 December, visited on 6 April 2015, http://cis-india.org/internet-governance/blog/invisible-censorship

14 Rishabh Dara (2012): "Intermediary Liability in India: Chilling Effects on Free Expression on the Internet," The Centre for Internet and Society, 27 April, visited on 6 April 2015, http://cis-india.org/internet-governance/chilling-effects-on-free-expres... .

15 Rule 75, Copyright Rules, 2013.

16 The Draft Counter Proposal is available at http://cis-india.org/internet-governance/counter-proposal-by-cis-draft-i...

17 According to Facebook's transparency report, there were 4,599 requests in the first half of 2014, followed by 5,473 requests in the latter half. Available at https://govtrequests.facebook. com/country/India/2014-H2/ also see Google's transparency report available at http: //www.google. com/transparencyreport/removals/government/IN/?hl=en and Twitter's report, available at https:// transparency.twitter.com/country/in

18 Surabhi Agarwal (2015): "Transparency Reports of Internet Companies are Skewed: Gulashan Rai," Business Standard, 31 March, viewed on 5 April 2015, http://www.business-standard.com/article/current-affairs/transparency-re... .

19 http://sflc.in/deity-says-2341-urls-were-blocked-in-2014-refuses-to-reve...

20 "32 Websites Go Blank," The Hindu, 1 January 2015, viewed on 6 April 2015, http://www.thehindu.com/news/national/now-modi-govt-blocks-32-websites/a...

21 Pranesh Prakash (2012): "Analysing Latest List of Blocked Sites (Communalism and Rioting Edition)," 22 August, viewed on 6 April 2015, http://cis-india.org/internet-governance/blog/analysing-blocked-sites-ri... . Also, see Part II of the same series at http://cis-india.org/internet-governance/analyzing-the-latest-list-of-bl... and analysis of blocking in February 2013, at http://cis-india.org/internet-governance/blog/analyzing-latest-list-of-b...

22 Gautam Bhatia (2015): "The Supreme Court's IT Act Judgment, and Secret Blocking," Indian Constitutional Law and Philosophy, 25 March, viewed on 6 April 2015, https://indconlawphil.wordpress.com/2015/03/25/the-supreme-courts-it-act...

23 Apar Gupta (2015): "But What about Section 69A?," Indian Express, 27 March, viewed on 5 April 2015, http://indianexpress. com/article/opinion/ columns/but-what-about-section-69a/

24 Pranesh Prakash (2011): DIT's Response to RTI on Website Blocking, The Centre for Internet and Society, 7 April, viewed on 6 April 2015, http://cis-india.org/internet-governance/blog/rti-response-dit-blocking ). Also see http://cis-india.org/internet-governance/blog/analysis-dit-response-2nd-... and http://cis-india.org/internet-governance/resources/reply-to-rti-applicat...

25 http://sflc.in/wp-content/uploads/2015/04/RTI-blocking-final-reply-from-...

The views and opinions expressed on this page are those of their individual authors. Unless the opposite is explicitly stated, or unless the opposite may be reasonably inferred, CIS does not subscribe to these views and opinions which belong to their individual authors. CIS does not accept any responsibility, legal or otherwise, for the views and opinions of these individual authors. For an official statement from CIS on a particular issue, please contact us directly.