The Information Technology Rules and the Bombay High Court’s Stay Order

The Bombay High Court’s order of 14th August 2021, which stayed the application of Rule 9 of the Information Technology Rules 2021, repays careful study. The Leaflet legal news portal and the journalist Nikhil Wagle had filed challenges to the Rules as a whole, and by way of interim relief, had pleaded for a stay on Rules 7, 9, 14, and 16. While the Court declined to grant a stay on Rules 7, 12, and 14, it did agree to stay Rule 9, which requires online publishers to mandatorily comply with a “Code of Ethics” (as set out), or face prosecution. The “Code of Ethics”, in turn, is contained in an appendix to the Rules, and includes compliance with the Press Council of India norms as well as the Programme Code under the Cable Television Networks (Regulation) Act.

As a perusal of the Gazette Notification reveals, the 2021 Rules are purported to be passed under the authority of Sections 87(2)(z) and (zg) of the IT Act 2000. Section 87 of that Act delegates rule-making power to the government, and 87(2)(z) specifically authorises the making of rules for the blocking of online content, while 87(2)(zg) does so for regulating intermediary liability. The Division Bench of Datta CJ and Kulkarni J held that Rule 9 was prima facie ultra vires Sections 87(2)(z) and (zg) of the IT Act, as, first, it envisaged penalties and consequences under entirely different statutory regimes (the Press Council Act and the Cable Television Networks Regulation Act) (paragraph 25); secondly, that the Information Technology Act – as a whole – did not intend to censor online content (except with respect to blocking under S. 69A) (paragraph 26); thirdly, Press Council norms under the PCI Act were themselves clearly stated to be limited to moral imperatives, and not legal ones. Under Rule 14(5) of the IT Rules, however, breach of the Code of Ethics (which included the PCI Norms, as pointed out above) could result in serious sanction, such as removal or deletion of content (paragraph 27); that if the Programme Code from the Cable Television Network Regulation Act was applied to publishers and curators of online content, it would lead to unconstitutional results, such as precluding them from criticising public figures (paragraph 28); and that consequently, the operation of Rule 9 would lead to a chilling effect upon online publishers, with the risk of being hauled up for punishment “should the inter-departmental committee be not in favour of criticism of any public figure.” (paragraph 29) Thus:

As it is, the constant fear of being hauled up for contravention of the Code of Ethics is a distinct possibility now. People would be starved of the liberty of thought and feel suffocated to exercise their right of freedom of speech and expression, if they are made to live in present times of content regulation on the internet with the Code of Ethics hanging over their head as the Sword of Damocles. This regime would run clearly contrary to the well-recognized Constitutional ethos and principles. (paragraph 30)

Thus, on the combined basis of ultra vires and a violation of Article 19(1)(a), the High Court held that the case for a stay had been made out (paragraph 32).

There are two brief comments that I want to make on this well-reasoned stay order. The first is that whether or not the learned Justices so intended, the reasoning in the stay order applies directly to the Cable Television Network Regulation Act and its Programme Code as well. It would be illogical to hold that the provisions of the Programme Code have a chilling effect on the freedom of speech when applied to publishers or online content, but do not when applied to television channels. This is especially true as the original – albeit flawed – distinction in the K.A. Abbas judgment between written content and audio-visual content, with the latter being held to a stricter threshold, obviously does not hold when you are comparing audio-visual content with online content. Datta CJ and Kulkarni J were entirely right to note that the provisions of the Programme Code, as applied in the context of the IT Rules, are prima facie unconstitutional, and for exactly the same reasons, it is submitted that they are unconstitutional per se (note that violations of the Programme Code are also – like in the case of the IT Rules – adjudicated in the first instance by a government committee). Thus, perhaps the time for a court to squarely examine them is at hand!

The second point is that the Court declined to grant a stay on Rule 16 of the IT Rules, which provide for blocking. Inter alia the Court observed that this provision was in line with S. 69A(1) of the IT Act, which authorised the blocking of online content where it was “necessary or expedient” to do so in the interests of any of the eight sub-clauses set out under Article 19(2) of the Constitution. Admittedly, the constitutionality of S. 69A had been upheld in the Shreya Singhal judgment, and consequently, there was not much the High Court could have done here even if had been so inclined. This only reveals – yet again – however, how unsatisfactory the Shreya Singhal judgment is on S. 69A. To start with, recall that the word “expediency” too broad and vague as a justification, while also falling outside of the proportionality standard, for the restriction of fundamental rights. Indeed, there is something ironic about a statute that allows for constitutional rights to be restricted when “expedient”; one might wonder what is left of the content of a right when it can be curtailed whenever the government feels it expedient. Furthermore, and bracketing this for the moment, it is submitted that Shreya Singhal erred in upholding a copy-paste job from Article 19(2) of the Constitution to the S. 69A. The point of Article 19(2) is to provide a set of standards that allow courts to judicially review speech-restricting laws. The point of Article 19(2) is not to provide a template for legislation that essentially takes the form: “the government may restrict rights when it is constitutionally correct to do so.” Under any standard, this would amount to impermissibly excessive delegation on part of Parliament to the Executive – because it is Parliament’s task to prescribe, with specificity, the circumstances under which the Executive is authorised to restrict rights, and to what extent.

As pointed out above, this is not intended to be a criticism of the Bombay High Court order, as the Court was, of course, bound by Shreya Singhal. It is interesting to note, however, that the stay order reveals that the challenge to the IT Rules 2021 is yielding both a new set of issues around online free speech for courts to consider, as well as demonstrating old pathologies anew. It will thus be important to track the further progress of this case through the judicial hierarchy.

Intermediary Guidelines and the Digital Public Sphere: Automated Filtering

In a previous post (here), I had argued that focusing on intermediary liability to regulate social media companies would likely weaken free speech and further entrench the power of large technology companies. However, with the adoption of the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 (“Intermediary Guidelines”), the government has adopted a liability centred approach – so let us take stock.

As the name suggests, the Intermediary Guidelines traverse well beyond the regulation of online intermediaries and raise several issues as discussed by the IFF (here and here) and the SFLC (here). The Guidelines are currently subject to a comprehensive legal challenge before the High Court of Kerala. At their core, the Guidelines change the obligations imposed on online intermediaries in ways that may fundamentally alter the public sphere. This post focusses on the obligation of social media companies to use ‘automated filtering mechanisms’ to weed out ‘illegal’ content.      

The Context

Content on the internet typically concerns three parties: (1) ‘content originators’ who post or share content; (2) online intermediaries who provide a platform for users to share content on; and (3) ‘content recipients’ who view the content posted by content originators. Conceptually, it is important to distinguish between websites that post their own content (a news organisations website), and online intermediaries that merely provide a platform (WordPress, which allows bloggers and readers to interact but has nothing to do with the content of individual blogs). When (inevitably) the material posted by a content originator violates the law, it raises the question of whether the online intermediary should be held secondarily liable for its role in spreading the illegal content.

Section 79(1) of the IT Act states that an online intermediary shall not be liable for this facilitative role. This ensures that intermediaries do not have to scrutinise each user post for potential illegality/liability that may be foisted on them, ensuring the free flow of user content on platforms such as Facebook and Twitter. However, given the wide reach of the internet, governments also typically want intermediaries to act as gatekeepers against some of the worst forms of content. As a result, in India, the immunity offered to intermediaries by Section 79(1) requires intermediaries to satisfy certain obligations. Crucially, under Section 79(2)(c), online intermediary must “observe due diligence”. This is where the Intermediary Guidelines become relevant, as Rules 3 and 4 of the Guidelines prescribe the ‘due diligence’ obligations of online intermediaries. In other words, if intermediaries want to avail of the legal immunity for any illegal content on their platforms, they must satisfy the conditions of Rule 3 and 4 of the Intermediary Guidelines.   

Scheme of the Guidelines

Rules 3 and 4 of the Intermediary Guidelines impose a system of differentiated obligations for ordinary online intermediaries and ‘Significant Social Media Intermediaries’ (“SSMIs”). As expected, the latter have significantly more onerous responsibilities to fulfill to avail of legal immunity. While this post focuses solely on SSMIs, it is worth spending a moment on this distinction. The Guidelines define an SSMI as an intermediary which: (i) has more “registered users in India” than a limit set by the Union Government; and (ii) primarily or solely enables online interactions between users and allows them to access, upload, and share content. The Union Government has since stated that the threshold for an SSMI is 5,000,000 “registered users in India”.

As a practical issue, it is unclear exactly how “registered users in India” will be calculated. Geographic locations are easily spoofed on the internet and the requirement is bound to create a few borderline cases involving some arithmetic gymnastics to avoid being classified as a SSMI. To add to this, platforms such as Signal and Telegram do not release country specific userbase data.

Taking a step back, it is also worth examining the trend to impose additional monitoring and filtering responsibilities on large social media companies. The increased burden on SSMIs to moderate their platforms is a tacit admission that these platforms form the bedrock of public discourse or a ‘digital public sphere’. In addition to cat videos, these platforms facilitate communication between citizens and the political centre of self-governance. Thus, these platforms aid democratic legitimating by allowing citizens to participate and shape public opinion. Suspending a citizen’s social media account is the modern-day equivalent of gagging them at a town hall. Simply put, SSMIs perform a very public function in our democracy and thus their regulation must balance the government interests in preventing harm in the public sphere (e.g., defamatory, or presumptively illegal content), with the free expression rights of internet users (included here is the right to access/receive information).  

Proactive Filtering Requirement

Rule 4 sets out the “due diligence” obligations of SSMIs. Rule 4(4) states that an SSMI:

shall endeavour to deploy technology-based measures, including automated tools or other mechanisms to proactively identify information that depicts any act or simulation in any form depicting rape, child sexual abuse or conduct, whether explicit or implicit, or any information which is exactly identical in content to information that has previously been removed or access to which has been disabled on the computer resource of such intermediary under clause (d) of sub-rule (1) of rule 3, and shall display a notice to any user attempting to access such information stating that such information has been identified by the intermediary under the categories referred to in this sub-rule.

The obligation in Rule 4(4) thus creates a best efforts obligation on SSMIs to utilise automated tools to proactively identify two classes of content. First, content that explicitly or implicitly depicts rape or child sexual abuse, and second, content that is exactly identical to content that has been taken down pursuant to Rule 3(1)(d). Under Rule 3(1)(d), content may be taken down pursuant to a court order or a government order (as noted here, there are virtually no limits on what the government may block). Although the provision does not expressly require an intermediary to block access to a page identified as hosting such content, absent a clear government clarification that users will still be permitted to view the flagged content, we may presume that the intermediary’s notice flagging the content will replace the web-page’s actual content – effectively blocking the content. (Sidebar | As takedowns under court or government orders themselves are haphazard and regularly not disclosed to the public, it is unclear what “categories” the user-facing notice will use, drastically reducing its usefulness as a transparency aid.)

Overcompliance and Monitoring

Rule 4(4) provides intermediaries a choice, they must either monitor their platforms for select types of illegal content and replace it with red-flag, or alternatively, they can forgo the legal immunity offered by Section 79(1) and risk being held secondarily liable for facilitating the spread of illegal content. Despite the musings of certain commentators, the impact of giving intermediaries such a choice should be obvious. As Jack Balkin notes, this approach ‘raises problems of collateral censorship and digital prior restraint.’ In other words, conditioning legal immunity on the effectiveness of a platform’s monitoring and takedown of illegal content is bound to make platforms overly cautious. They are likely to takedown content that could potentially be illegal but may not necessarily be so if adjudicated upon. As Balkin notes, this phenomenon is inherent to systems of intermediary liability as intermediaries have no incentives to protect the speech of unknown third-party users over their own coffers.

The result is that intermediaries may restrict a user’s post purely to avail the legal immunity provided by Section 79(1) – because they do not want to take the risk that if the post is later found to be illegal and they had failed to take it down, they could be held liable for disseminating it. Crucially, an intermediary may takedown/block speech that a user could tenably claim is constitutionally protected speech, but the user’s claim does not sufficiently alter the risk calculus of the intermediary, which operates under the outsized threat of liability. This problem is aggravated by the oddity that users do not have free speech rights vis-à-vis online platforms (private companies) yet these platforms undeniably constitute part of the digital public sphere.

To avoid this very problem, the European Union expressly prohibits imposing a ‘general monitoring obligation’ on intermediaries to weed out illegal activity on their platforms (see Art. 15 E-Commerce Directive). In the context of requiring intermediaries to weed out copyright infringing material from their platforms, the European Court of Justice ruled as far back as 2012 that a general monitoring obligation was incompatible with the free speech rights of users as it may lead to legal content being caught in the dragnet of a web-filter. The court held that a general monitoring obligation exists where intermediaries had to install filters that: (i) were targeted at user content; (ii) applied indiscriminately to all users; (iii) were installed as a preventive measure; (iv) at the sole expense of the intermediary; and (v) for an unlimited period of time. Needless to say, Rule 4(4) ticks all the boxes of an unlawful general monitoring obligation.

Using Automated Filters

The use of automated systems is not new. Most large social media platforms already voluntarily use a host of automated tools to filter content. However, these are by no means perfect. Between Facebook’s takedown of iconic war photography, to Twitter’s suspension of an Indian lawyer’s account over the use of August Landmesser’s photo defying the Nazi salute – the mistakes of automated systems are well documented. Put simply, at their present state of development, automated systems are poor at evaluating cultural context, irony, or the critical analysis necessary to discern between illegal and legal content. Now consider the types of content Rule 4(4) requires such automated tools to block, ‘explicit or implicit depictions of rape and child sexual abuse’. Automated systems are unlikely to accurately determine the various layers of nuance associated with such content which may range from actual footage of child sex abuse to a critical commentary on sexual violence.

This issue is exponentially aggravated when the second target for automated filters is considered. Content may be taken down under Rule 3(1)(d) for a host of legal and legal but unconstitutional reasons. A court may pass an order to takedown defamatory content under Rule 3(1)(d). The Government may also pass an overbroad order blocking large swathes of content under Rule 3(1)(d). The effect is to create an ever-growing database of illegal content (infinitely more complex than even depictions of rape or child sex abuse) that automated systems must continually weed out. Furthermore, under Rule 3(1)(d) specific URLs are restricted. If an automated system is unable to discern exactly what content on that web-page was illegal, there exists a significant chance of collateral censorship as all content on that originally blocked page may be flagged as illegal and automatically blocked going forward. Finally, there are no safeguards to ensure that once content is no longer restricted under Rule 3(1)(b) (if a court modifies or the government withdraws an order), automated systems are immediately turned off.

Although Rule 4(4) uses the term “proactively”, it is unclear whether SSMI’s will be obligated to restrict content prior to publication even. A filtering system that restricts content prior to publication fundamentally privileges the interests in favour of restraining speech. In other words, it treats all speech identified by the filter as presumptively illegal, contrary to settled free speech doctrine. Lastly, even if the content is published, a user must be given the right to contest the application of the filter before it is taken down. While Rule 4(8) of the Intermediary Guidelines does postulate an ex-ante dispute resolution mechanism, it does not apply to content taken down pursuant to the automatic filtering under Rule 4(4). This is amplified when the relevance of the content is time sensitive, and no content-recipient has been identifiably injured. Famously, in Shreya Singhal, the Supreme Court struck down a requirement that intermediaries takedown content at the behest of a private complaint absent judicial adjudication. Where this leaves a requirement that intermediaries takedown content even prior to a private complaint will have to be seen.

Privacy

The use of automated filters also raises issues surrounding user privacy. Imposing a general monitoring obligation on intermediaries to constantly weed out “illegal” content requires them to conduct an ongoing analysis on all user content posted on their platforms. This exercise is likely to lead to a large dataset that links user profiles to any “illegal” content posted by them. Consider a user profile which repeatedly posted content that violated a ‘depictions of rape’ filter or ‘farmer protests’ filter – the intermediary would be in possession of this information which may: (i) be requisitioned by the government; or (ii) be hacked or leaked and made available to the public devoid of any context.

Parchment Barriers

The Union Government is not unaware of these harms, and Rule 4(4) includes three safeguards aimed at preventing these issues. First, the measures taken by intermediaries must be “proportionate” and must consider the ‘interests of free speech and privacy of users.’ Second, intermediaries shall implement “mechanisms of human oversight … including a periodic review of any automated tools deployed”. Lastly, the intermediary “shall evaluate the automated tools having regard to the accuracy and fairness of such tools, the propensity for bias and discrimination in such tools, and the impact on privacy”.

On paper, these are excellent considerations to have when deciding whether to use automated filtering. However, as regulatory text, it lacks the necessary specificity or enforceability to reign in the dangers of automated systems. None of the three safeguards provide any indication as to when an intermediary is in breach of them. How much oversight and review are necessary to maintain immunity? What is the level of efficacy the automated systems must achieve and what is a tolerable error rate? What forms of data collection for the purposes of such systems are impermissible? Additionally, the Guidelines and the IT Act fail to provide any form of regulatory oversight. There is no dedicated body to police these incredibly complicated systems being unleashed on Indian users, meaning that in the real world – compliance will be almost impossible to ensure on an ongoing basis.

Automated systems are not unregulatable. There exist several real world proposals on how to monitor such systems such as vetting the source code, black boxes, and perhaps most promisingly – ‘tinkering’. This involves systematically testing how such filters respond to sample content on an ongoing basis. However, such solutions require a regulatory framework and committed agencies. The safeguards set out in Rule 4(4) are more akin to parchment barriers – laudable on paper but easily circumventable by both intermediaries and the government in practice.

The so-called safeguards also raise the spectre of a more sinister threat that is now well-established in India’s legal system: that of ‘discriminatory legalism’. Simply put, the legal uncertainty and lack of a dedicated and independent agency to oversee SSMI filtering opens the door for the government to selectively prosecute SSMIs for either violating the filtering obligation or violating the safeguards. For example, if the SSMI does not takedown content unfavourable to the ruling party, a violation of the web-filtering may be alleged, and if the SSMI takes down content favourable to the ruling party, a violation of the safeguards may be alleged.  

Conclusion

This is the first post of a series examining the impact of the new Intermediary Guidelines on the digital public sphere in India. Requiring SSMIs to take an increased role in content moderation of public discourse is not an unreasonable demand. The deregulatory atmosphere in which online intermediaries emerged means that private companies now perform the core democratic function of facilitating public discourse and regulation must recognise this. The large volume of content on these platforms also means that the use of automated filters is inevitable.

However, this set of posts critically examines whether the form of regulation adopted by the Intermediary Guidelines is the best way of ensuring a vibrant digital public sphere. Making intermediaries’ legal immunity conditional on them ceaselessly monitoring content fundamentally alters their risk calculus towards content takedown. The lack of well-conceived regulatory framework to address the risks of automated filtering by social media companies leaves Indian citizens more at the mercy of large tech companies and government interference than ever. A diametrically opposite, but equally feasible approach suggested by Martin Husovec is to fine intermediaries for over-zealously taking down content. This incentivises intermediaries to invest in improving their content removal systems, as opposed to India’s approach which, at best, allows the social cost of lost speech to go unaddressed – and at worst, incentivises intermediaries to takedown more content than ever.

“Fake News” and the Constitution

As millions of migrant workers made their way from India’s cities back to their villages after the government announced a nation-wide ‘lockdown’, the Solicitor General of India informed the Supreme Court that the exodus was caused by “some fake/misleading news and social media” and sought a direction to prevent “fake and inaccurate reporting” (here). In Maharashtra, an order was passed under Section 144 of the Code of Criminal Procedure prohibiting the dissemination of information on social media that was ‘incorrect or distorted facts’ (here). And Kashmir’s new ‘Media Policy – 2020’ states that “Any fake news or any news inciting hatred or disturbing communal harmony shall be proceeded against under IPC/Cyber Laws” (here).

Phrases such as “incitement” and even “disturbing communal harmony” have a long and well-documented use in Indian law (for better or for worse). However, the above narrated incidents demonstrate a recent trend by the Indian government to try and restrict speech on the ground that it constitutes “fake news” (I use speech in the broadest possible term to include the press, broadcasting and online media). India is not alone, countries such as Singapore and Indonesia have introduced full-blown legislation to restrict “fake news”. In this post, I begin by noting that the term “fake news” suffers from several definitional hurdles that point to deeper structural problems in our media eco-system. I argue that there are several very good reasons why we may want to restrict some forms of misinformation. However, any restriction imposed on speech must comply with the constitutional safeguards set out in Articles 19(1)(a) and 19(2). Examining “fake news” restrictions against the concepts of vagueness, overbreadth and a disproportionate chilling effect, I argue that restrictions on “fake new” that are narrowly tailored enough to be constitutionally compliant are unlikely to be effective in combatting the social harms we associate with “fake news”. I conclude by advocating a heterogeneous approach to combat the issue of “fake news”.

A few caveats. First, because India does not yet have a “fake news” legislation, my analysis is necessarily in the abstract (even the Kashmir policy ultimately relies on provisions of the Indian Penal Code for prosecution). This post seeks to evaluate the consequences of restricting “fake news” as a category of speech and I accept that any restrictions imposed by the government may be more nuanced than a blanket restriction on “fake news” (although the signs are not promising). Second, there is a separate but cognate conversation to be had about the role of internet intermediaries in facilitating and restricting “fake news” that is worthy of a separate post and I have not addressed the issue here for the sake of brevity.

Protected Speech and its Limits

Before beginning it pays to recap a few important aspects of free speech regulation in India. While Article 19(1)(a) guarantees citizens the freedom of speech, Article 19(2) allows for “reasonable restrictions” in the interests of inter alia: (i) the sovereignty/integrity of India; (ii) the security of the State; (iii) public order; (iv) decency or morality; (v) defamation; or (vi) incitement to an offence. As we can see, speech in India can be restricted because of its consequences, that it may lead to violence, but also because of the speech’s content – that the meaning conveyed is deemed legally objectionable. The State evidently has an interest in restricting speech that directly leads to violence. However, in the case of obscenity laws or defamation, speech is restricted because of value judgements by the State. Obscene speech does not lead to violence, but the State believes that it leads to an erosion of public morality.

Any restriction on speech must have a proximate connection with a specific head set out in Article 19(2). The government cannot restrict speech merely in the ‘public interest’, or because it is ‘false’, neither of which are heads under Article 19(2). Therefore, if the government wanted to restrict “fake news” it would need to prove that “fake news” either caused harm because of its content (defamation, decency or morality) or that it was inciteful leading to violent consequences (public order, incitement to an offence). Lastly, there is a long line of cases noting that the ‘proximate connection’ means a real and imminent risk of harm arising from the speech and not vague speculation about possible future harms.

Defining “Fake News”

“Fake news” is a term bandied about very loosely nowadays which has resulted in everybody thinking there is consensus about the phenomenon being referred to, but very little certainty as to what content is “fake news” and what content is not. The term has been applied to satire, propaganda, biased reporting, sponsored or promoted content, factually incorrect reporting, entirely fabricated stories, or simply inconvenient truths. The term does not clarify whether it applies to private communications (WhatsApp chats), social media (Facebook), online media (an online-only news organisation) or even traditional print media. In a post Donald Trump era, the term also necessarily carries a derogatory component that is often independent of an objective evaluation of the actual content (Habgood-Coote refers to this as an ‘epistemic slur’). The flip side of this issue is that there is very little certainty about who a “journalist” is today, with citizens receiving news from a wide variety of sources.

The term “fake news” therefore refers to a heterogeneous field of content, some of which have a diverse set of underlying problems. This becomes immediately apparent when we look at the words we used to use to describe this type of content before we began using the umbrella term “fake news”. Inaccurate, false, misleading, biased, sensationalist, propaganda and advertisement are just some of the words to describe what we now call “fake news”. Using an exact term to identify the issue with a piece of content allows us to create targeted and meaningful solutions. For example, the type of regulation needed to regulate factual inaccuracies in a newspaper article is very different from the type of regulation needed to ensure paid advertising can be distinguished from news stories. We need to stop using the umbrella term of “fake news” and begin accurately labelling the specific harms caused by the speech in question.

From a free speech perspective, it is important to note that the distinction between “fake” and “real” is a politically contested one. The more polarised a society, the less likely it is that different parts of a population experience the same political reality. This makes legally regulating the fake/real distinction problematic, as seen in Singapore where the government sent orders to Facebook to “correct” individual posts. The posts alleged that the Singaporean government had illegally influenced investment companies, suppressed whistle-blowers, and rigged elections (here). This is emblematic of how governments can use a restriction on “fake news” to restrict a broad range of criticism and plenty has been written about the Indian government’s efforts to reshape the narrative in Kashmir (here). At the end of the day, governments place a pre-eminent value on self-preservation coupled with a bias towards their own political ideologies, and the ability to determine what is “fake” and what is “real” goes a long way towards silencing opposing viewpoints and homogenising political thought.

That said, there exist more nuanced definitions of “fake news” and a blanket ban on “fake news” is unlikely. For example, the European Union (in non-binding documents) doesn’t use the term “fake news” at all, rather it defines “disinformation” as ‘verifiably false or misleading information which cumulatively is created, presented and disseminated for economic gain or to intentionally deceive the public and may cause public harm intended as threats to democratic, political and policymaking process as well as public goods’. By requiring an evaluation of the intent behind the creation of the information and a (rather soft) the requirement for actual harm the European definitions seems to be less of a burden on free speech. But this also means it is less effective at curbing the spread of misinformation. Users may spread disinformation legitimately believing it to be true, and how does one assess when a ‘policymaking’ process is harmed? It also does not regulate other harms we associate with “fake news” such as media bias. It is worth keeping these considerations in mind as we consider the harms arising from “fake news” or “disinformation”.

The Argument for Regulating “Fake News”

There are two primary reasons why a government may legitimately wish to regulate “fake news”. First, misinformation or fabricated stories may directly result in violence, either through information about specific individuals or more broadly stoking pre-existing fissures in society (e.g. race or religion). India has already witnessed a string of violent incidents that investigative authorities have noted were either caused by or aggravated due to the spread of disinformation. Second, disinformation can interfere with the electoral process by misinforming voters about candidates’ political opinions, track records, previous misdeeds, and positions on important social issues. An uninformed voter is unable to vote for a candidate that best represents their interests, directly undermining the legitimacy of the electoral process and the resultant government. To make matters worse, disinformation is likely to disproportionately effect voters who do not have access to multiple sources of information (the poor and marginalised).

As Ari Waldman notes, in the long run disinformation can fundamentally alter public discourse by creating false equivalencies, particularly where political leaders spread disinformation. If a factually incorrect statement by a political leader must be accorded equal weight as the truth in the name of journalistic neutrality, a society may waste valuable public time and energy debating false stories (e.g. see the amount of time the U.S. media spends merely correcting President Trump’s demonstrably false statements). Over time, sustained bias or propagandistic reporting can harden political bias, causing citizens to select media sources that merely confirm their existing notions of true or false, further increasing polarisation.

To be clear, traditional justifications for the freedom of speech such as Mill’s argument that more debate over ideas ultimately lead to the truth (culminating with Justice Holmes famous ‘marketplace of ideas’ analogy) are not strictly applicable to “fake news”. To argue that we should produce even more “true news” (counter-speech) rather than restrict “fake news” is problematic for two reasons. First, it doesn’t work and inaction can lead to the type of cyclical harms referred to above (a recent MIT study found that false stories diffused further and faster than true stories on Twitter in all categories of information  (here)– in other words, the truth does not rise to the top).

More fundamentally, Mill’s theory and subsequent adaptations are premised on having more ideas and counter-ideas, not facts and “counter-facts”. Facts and opinions are very distinct in their nature but also share an important relationship when it comes to free speech. While there may be an infinite number of ideas or opinions about a fact, a fact is singular and objective. For example, whether there are Chinese soldiers inside what India considers its territory is a question of fact. The soldiers are either there, or they are not. What India should about the Chinese soldiers and its own territorial claims is a question of opinion with many possible answers. However, it is also important to acknowledge that “factual truth informs political thought” – in other words people’s opinions have a factual basis. For people to have meaningful opinions about the world, and for those opinions to interact with each other in meaningful debate, there must be an agreed upon factual basis. It makes sense not to restrict ideas and opinions because they open new avenues of thinking, however an agreed upon baseline of facts is fundamental to an informed debate about ideas. Returning to our example, how can we debate whether the government’s foreign policy was a success if we do not know whether the Chinese soldiers were in Indian territory or not? Therefore, a freedom of speech justification created to ensure free debate may actually favour some restrictions on “fake news” by ensuring a common baseline of facts.

Legally Restricting “Fake News”

Recall that under the constitutional scheme, all speech is free other than that which the government restricts because it is has a proximate nexus with the specific harms identified by the heads of Article 19(2) (public order, defamation etc.). Although the Indian Supreme Court has never explicitly dealt with the question of whether factually false statements are protected under the Constitution (the U.S. Supreme Court in Alvarez explicitly held such statements to be protected under the U.S. Constitution) factual inaccuracy is not a ground to restrict free speech under Article 19(2). Therefore, it is reasonable to assume that if the government wanted to restrict speech it classified as “fake news”, it would have to argue that the speech was either defamatory, would lead to a breach of public order, or amounted to an incitement to an offence. What amounts to a valid restriction under these heads of 19(2) have been widely discussed elsewhere on this blog and I do not intend to rehash them. Rather I argue that restrictions on “fake news” are likely to fall foul of two principles: (i) vagueness, (ii) overbreadth leading to a disproportionate chilling effect on speech.

Vagueness: A law is unconstitutionally vague if ordinary citizens cannot determine whether they have broken it or not. Imagine a law which restricted citizens from “honking too much” at traffic signals. How does a citizen determine how much is “too much”? The law does not provide a fair warning to citizens as to whether their actions will break the law or not. Vague laws also grant officials a large amount of discretion as to when a law has been breached or not (would you trust a police-officer to tell you when you are honking too much?).

Restricting speech on the ground that it is “fake news” would likely lead to widespread confusion about what kind of speech the government was restricting. Recall the broad range of content that comes under the banner of “fake news” – from satire to inaccurate reporting to fabricated stories. More nuanced definitions (such as that proposed by the European Union) may remedy this issue but are still unwieldy. The line between advocacy and deception is often imperceptible, with compelling arguments often cherry-picking or manipulating facts. Further, terms such as ‘threats to the democratic, political and policymaking process’ are so wide that they substantially increase the risk of the government selectively prosecuting speakers with unfavourable opinions. Kashmir is a living example of where free speech has been left entirely to the whims of the executive, and it has led to the denial of internet services, the arbitrary blocking of websites, and the persecution of journalists.

Overbreadth: A restriction on speech is “overbroad” when it restricts both the speech that the government can legally restrict (e.g. hate speech) but also goes on to restrict speech other speech that is constitutionally protected (e.g. dissent). In Shreya Singhal, the Supreme Court was called to adjudicate upon the constitutionality of Section 66A of the Information Technology Act, which criminalised speech that was “grossly offensive”. The court noted that any citizen may advocate a view on governmental, literary, or scientific issues that may be unpalatable or even “grossly offensive” to other citizens. While some speech may be justifiably restricted, not all speech that was “grossly offensive” rose to the level where there was an imminent risk to public order or incitement. This meant that, by using the term “grossly offensive” Section 66A also ultimately criminalised constitutionally protected speech. The court observed:

Section 66-A purports to authorise the imposition of restrictions on the fundamental right contained in Article 19(1)(a) in language wide enough to cover restrictions both within and without the limits of constitutionally permissible legislative action […] It must therefore, be held to be wholly unconstitutional and void.”

The term “grossly offensive” was so broad that it restricted some speech justifiably, but also other speech unjustifiably. In Shreya Singhal the court distinguished between discussion, advocacy, and incitement. It held that the first two formed the heart of constitutionally protected speech while inciteful speech may be legally restricted. Simply put, restrictions on speech that restrict inciteful speech and are broad enough to go on and restrict discussion and advocacy are unconstitutional.

If the government were to restrict “fake news” (a very broad category of speech), it would also likely restrict a vast amount of constitutionally protected discussion and advocacy. At its bluntest, satire, parody, and anti-government reporting could fall under the ambit of “fake news”. More critically, a restriction on “fake news” would create the risk of liability for journalists and media houses, that a single inaccurate factual assertion could lead to censorship or punitive action. For example, Kashmir’s new ‘Media Policy’ flat out dis-empanels journalists for publishing “fake news”. A news-report that exposed governmental overreach but contained a minor factual error or irregularity could be termed “fake news”, robbing readers of valuable information and robbing journalists of their livelihood.

Where restrictions on speech are vague, overbroad, and punitive, they create a chilling effect on speakers. Content creators, journalists, publishers, media houses and ordinary citizen are likely to ‘err on the side of caution’ and simply not speak rather than run the risk of their speech being adjudged as a crime. To avoid this chilling effect, the law of defamation incorporates what is known as the “actual malice” standard. It states that unless a speaker’s statement evidences ‘actual malice or a reckless disregard for the truth’, the statement cannot be considered defamatory. The logic is simple, with the vast amount of free-flowing debate in society a few inaccuracies are bound to crop up. Rather than aggressively prosecute these minor inaccuracies and frighten all other speakers, the law states that where these inaccuracies do not possess any ‘actual malice’ they are exempt from prosecution. In Rajagopal the Indian Supreme Court adopted the ‘actual malice’ standard in civil defamation cases and recently, the Madras High Court has adopted the standard in the context of criminal defamation (here). Defamation requires an alleged injury to reputation, something not all “fake news” is likely to cause. However, the rationale that minor inaccuracies should not lead to punitive action which has a chilling effect on speech should certainly apply, and any restrictions on “fake news” should incorporate the ‘actual malice’ standard.

The Government and the Truth

One last point may be made before concluding. In Alvarez, the U.S. Supreme Court struck down a statute which punished persons for falsely claiming to have been awarded military medals. In striking down the law, the court noted that that it was wary of the government claiming ‘broad censorial power to regulate falsehoods’, the mere existence of which would have a chilling effect on speech. To be clear, the government does regulate falsehoods in certain cases, most notably in the areas of defamation, consumer protection and fraud, and perjury. This is accepted because, the restrictions on speech are limited to the contexts where they are especially likely to cause harm and are actionable only where they cause actual harm to identifiable individuals. A restriction on “fake news” is much broader. A good news eco-system is more akin to a public good like a clean environment and “fake news” often attacks issues more than people – where it does attack people, the remedy of defamation is always open. Similarly, where the “fake news” is inciteful and capable of causing violence, India already possess a host of anachronistic laws that can be used to restrict and prosecute speech capable of causing violence. Without this proximate connection to real threats, regulating “fake news” creates the spectre of a government truth, something any democracy should be wary off.

Conclusion

The phenomenon often branded as “fake news” can lead to a diverse set of harms ranging from violence, damage to the electoral system and increased political polarisation. However, when examined in light of India’s freedom of speech jurisprudence, restricting “fake news” is constitutionally fraught. Above all, it is important to remember that the government cannot restrict speech merely because it is “false” or “inaccurate”. Briefly, any restriction on “fake news” would need to (i) show some real, imminent and identifiable harm; (ii) the harm would need to be a constitutionally recognised ground to restrict free speech under Article 19(2); (iii) consider questions of medium, time and impact of the “fake news”; (iv) incorporate the ‘actual malice’ standard; and (vi) be the least restrictive measure at the government’s disposal. This is not to say that it is impossible to draft legislation regulating “fake new” (several countries have already done it). There may be real value in pursuing narrow restrictions to reduce the effects of active disinformation campaigns during especially sensitive times (e.g. the run-up to elections). However, the phenomena of “fake news” rarely appears in such convenient forms and speech restrictions that are constitutionally compliant are unlikely to address the other diverse and systemic harms that “fake news” causes.

I began by noting that “fake news” is an amalgam term that houses several underlying issues we associate with our media-ecosystem. Understanding the underlying issues can lead to targeted solutions that bolster media literacy amongst the population and reduce the economic and political incentives associated with fabricated or propagandistic stories. Examples include ensuring a competitive media industry, disclosures requirements on social media sites about promoted content, investment in a truly independent state broadcaster, media literacy education in schools, and ultimately targeted legal interventions where constitutionally compliant speech restrictions may be meaningfully enforced. Merely sounding alarm bells at the rise of “fake news” opens the door to restrictions on speech and government censorship.

This post was largely in response to Kashmir’s Media Policy, which along with the Solicitor General’s statements represent a worrying trend by the government to adopt the clumsy but incredibly dangerous fake/real distinction to assert a dominant, government controlled narrative. In the future, I am sure there will arise more concrete examples that lend themselves to more detailed critique. The government’s position has long been that speech in Kashmir rejects the paradigm of the Indian State itself, representing a unique threat to the integrity of India. However, where the government seeks to preserve power through a stranglehold on truth itself, we have to ask ourselves whether the means employed to preserve the State have overridden what the State once stood for.

Notes From a Foreign Field: The Ugandan Constitutional Court on the Right to Protest [Guest Post]

[This is a Guest Post by Karan Gupta.]


In a society where policy brutality and clamp down on free speech is common, the Constitutional Court of Uganda recently affirmed a few commonsensical principles on free speech, the right to assemble and public order. On 26 March 2020, the Constitutional Court of Uganda declared Section 8 of the Public Order Management Act 2013 (POMA) unconstitutional (4-1 majority). Section 8, inter alia, conferred upon the Inspector General of Police (IGP), or any officer authorized by them, vast discretionary powers to: (i) Withhold permission to hold a public meeting or stop a public meeting where it is “held contrary to the Act”; (ii) Use force to disperse public meetings; and (iii) Impose criminal liability on organizers and participants of such public meeting. The Act defines a ‘public meeting’, empowers the IGP to regulate their conduct (S. 3), and requires every ‘organizer’ to give prior notice (at least three days prior and no more than fifteen days prior) of the proposed public meeting with details specified therein (S. 5). An unplanned, unscheduled and unintended public meeting is exempt (S. 7).

I explore, in seriatim, the constitutionally flawed approach, the progressive observations of the constitutional court, missed opportunities, and lessons for India.

 Preliminary Point: Constitution adjudication – in personam?

In a previous case, Muwanga Kivumbi v Attorney General, the Constitution Court had declared Section 32(2) of the Police Act 2006, which empowered the officer in charge of the police to pass an anticipatory order prohibiting the convening of an assembly/procession if there were “reasonable grounds for believing” that there would be a breach of peace, unconstitutional. The Court had held that the subjective and anticipatory power was prohibitory in nature (as compared to a regulatory power, which is permissible) and ultra vires Arts. 20(1) (Fundamental Rights are inherent and not granted by the state) and 29(1)(d) (Freedom of assembly and demonstration).

The challenge in HRNU lay in narrow confines. Art. 92 of the Constitution restricts the Parliament from passing a law which “alters” a decision of the Court “as between the parties to the decision or judgment”, thus barring the alteration of rights that have accrued to parties to a case vis-à-vis each other (in personam). The petitioners in HRNU highlighted that one petitioner in Muwanga was also a petitioner before the Court in HRNU, and that Section 8 of the POMA was para materia to Section 32(2) of the Police Act. Consequently, by enacting Section 8, the legislature had unconstitutionally attempted to alter the decision in Muwanga [p. 7]. Despite a broad challenge to the POMA on a myriad of constitutional provisions, the petitioners restricted their oral arguments to only Art. 92 and Section 8.

Justice Cheborion evaded the limited ambit of Art. 92 and the nuanced differences between the two provisions (Sections 32(2) and 8)) and held that Art. 92 also applies to decisions made in public interest, and not only in relation to parties to a previous litigation. This approach was adopted by two other judges [p. 42, 50, 69] and raises two concerns: First, this militates against the plain and ordinary meaning of Art. 92 and renders nugatory the latter part restricting its application to parties in a litigation. The decision in Muwanga (as well as the present case) concerned a constitutional matter on the ambit of constitutionally permissible police powers to control assemblies, demonstrations and peaceful protests. Such matters are, by their very nature, in rem proceedings – a declaration of invalidity does not operate only between parties, but to everyone. Art. 92 which seeks to protect rights that accrue to parties from a litigation (from contract law or property law for example), has no applicability in such cases. Furthermore, the legislature did not seek to alter the decision in Muwanga, but enact a new provision different from Section 32(2) of the Police Act.

Second, the Ugandan Parliament may alter the basis of decision in Muwanga by amending the provisions on which the decision turned [i.e. Arts. 20(1) and 29(1)(d)]. However, the Court held that the decision in Muwanga could only be altered, inter alia, by amending Art. 92 [p. 19]. This flows from its erroneous reading and application of Art. 92. The right approach is for the Court to, absent any constitutional amendment, employ in its assessment the broad constitutional principles laid down in Muwanga and other relevant constitutional provisions. However, the court restricted its assessment to whether Section 8 is an “incarnation” of Section 32(2) [p. 18]. Only two judges avoided this pitfall, though without adequate explanation [p. 54, 75], analyzed the entire Act and declared it unconstitutional [p. 61].

 Public interest, public order and free speech

Art. 29(1) guarantees to every person the freedom of speech and expression, assembly and demonstration, and association. While Art. 79(1) empowers the Parliament to enact laws for the maintenance of order, both provisions are silent on the permissible restrictions on fundamental rights. The answer is found in Art. 43 which stipulates that “no person shall prejudice the fundamental or other human rights and freedoms of others or the public interest”. According to Ugandan precedent, there is no justification to restrict or abrogate a fundamental right where the its exercise comports with the restrictions in Art. 43. Art. 43(2) clarifies that the term ‘public interest’ shall not permit any limitation “beyond what is acceptable and demonstrably justifiable in a free and democratic society”.

The scheme of fundamental rights chapter is significant for two reasons: First, barring the general restrictions in Art. 43 and a few other provisions which specify restrictions therein, there are no specified grounds to restrict fundamental rights as compared to other Constitutions; Second, under Art. 43, the principles of a free and democratic society are accorded primacy and any restriction must comport with this requirement. While the Parliament may legislate on the maintenance of public order (which Muwanga held is in public interest), any restriction on the freedom under Art. 29 must be demonstrably justifiable in a free and democratic society.

Viewed in the above context, the Court (three or more judges) made three significant observations:

 First, on the ambit and hierarchy of free speech protection, the Court held that speech, public processions and protests, irrespective of their nature, are entitled to equal protection (i.e. social, religious, political, economic, and so on) [p. 20, 57]. This is distinct from the preferred position doctrine which accords higher protection for political speech in American constitutional jurisprudence. The context which informed this analysis is granting equal protection to political assemblies and speech which are a common target of the political establishment in Uganda;

Second, on public order, the Court held that where a protest or public gathering is peaceful, “it does not matter that it may disruptive or even inconveniencing”. [p. 21] This is significant as any society committed to the freedom of speech and assembly recognizes that some disruption is no ground to restrict or deny the right. Beyond toleration, the celebration and protection of speech and assembly is linked to justice, equal concern and mutual respect of every individual. Recall here the judgment of the Madras High Court which affirmed that “public streets are the natural places for expression of opinions” (analyzed here); and

Third, the Court held that Section 8, in so far as it authorizes the police to prevent a public meeting, empowers them to impose a blanket ban and require prior permission for every gathering [p. 23, 25]. The Court held that this violates Arts. 29 and 43 and the state failed to demonstrate that the power conferred by Section 8 is both regulatory and acceptable and demonstrably justifiable in a free and democratic society. This is significant for three reasons:

  • The burden of proof to justify the restriction on a constitutionally guaranteed right falls on the state. However, by vesting in the police the vast discretion to prohibit or prevent the freedom of speech and expression, this burden stands reversed. Every individual is then required to justify to the police why the exercise of their constitutional right will not or does not impair public order or contravene any provision of POMA. Justice Elizabet rightly noted that there is a presumption that every assembly and exercise of free speech is peaceful [ 61] and a mere apprehension of violence does not constitute a sufficient basis to prevent or prohibit an assembly, gathering or protest;
  • The Court highlighted that no conditions were laid down for the exercise of the power [ 98], vesting in the police vast discretion to determine which public meetings may be prevented or forced to disperse. As noted by the American Supreme Court in Grayned v Rockford, such unbridled discretion is accompanied by the “attendant dangers of arbitrary and discriminatory application.” Where the police are an instrumentality of the state, the possibility of partisan politics to curb dissent and anti-establishment sentiments cannot be discounted. This is recognized by two judges in HRNU who documented the arbitrary exercise of power by the police to protect government interest and impose popular morality [p. 22, 100]; and
  • Flowing from above, the unregulated blanket discretion to prohibit assemblies does not comport with the requirements of the ‘constitutional yardstick’ that every restriction on a fundamental right must be necessary and The police must justify, in each specific instance, why the prohibition of an assembly is the least restrictive measure and proportionate to the possible harm sought to be prevented (this post discusses context-specificity and proportionality in the context of the internet shut-down judgment in Kashmir). The judgment, despite lacking in the explicit use of this yardstick, comports with this requirement.

Maintaining public order

How then must the State maintain public order? Justice Cheborian (with whom three other judges agree) answered this. He held that where the police anticipate a breach of peace, there is a positive obligation on the state to provide protection and police deployments and not prohibit the assembly [p. 17, 60, 90]. The Court held that the duty to maintain public order “cannot be discharged by prohibiting sections of the public from exercising their constitutionally guaranteed rights to demonstrate peacefully or hold public meetings of any nature.” This is sound as it reaffirms: (i) the holding in Muwanga that the state must provide channels and structures to ensure that legitimate protest “find voice”; and (ii) the principle that the failure of the state to provide adequate security cannot be a ground to deny people the freedom of speech and assembly. Furthermore, the Court noted that the state is also empowered to act in various situations (unlawful assemblies, riots, malicious damage against public order) by the Penal Code of Uganda [p. 33,89] and may, in accordance with the law, arrest or take appropriate action against any perpetrator [p. 23, 32].

Missed opportunities

 Despite the significant observations above, the Courts missed out on two opportunities:

First, only Justice Kenneth attempted to specify which values underlie a free and democratic societyinter alia, the acceptance and accommodation of a variety of cultural, religious and political beliefs and free political debate, human dignity and freedom of speech, association and movement [p. 90]. The Court could have laid down a comprehensive base for the protection of free speech and association as an values inhering in a free and democratic society; and

Second, despite the challenge in the petitions to numerous provisions of the Act, the Court examined the validity of only Section 8 (admittedly, only this was pressed by the petitioners). Only two judges examined the deeply inherent flaws in the entire Act to conclude that it was ultra vires a myriad of constitutional provisions. In the end, these draconian provisions were left standing and the Court concluded that guidelines must be framed for the exercise of powers under the Act [p. 33]

India and Section 144

Recall here that Section 144 of the Indian Code of Criminal Procedure 1973 confers wide discretionary powers upon executive magistrates to prohibit assemblies. This is regularly invoked on the basis of an apprehension that there would be a breach of peace or public order. The Constitution Court in Muwanga struck down a similar provision [Section 32(2) of the Police Act] on the ground that it was prohibitory in nature and reversed the burden of proof – every individual was required under it to justify why the exercise of the right to free speech and assemble would not cause a breach of public order. The Court held that this suppresses the “powerful tool” of peaceful assemblies and protests when a free and democratic society must encourage the “greatest possible freedom of expression.” This reasoning was reiterated in HRNU to strike down the wide power to impose blanket anticipatory bans on assemblies under Section 8 of the POMA.

The power under Section 144 [similar to Section 32(2) and Section 8] allows the imposition of anticipatory bans and is prohibitory in nature This falls foul the constitutional standards espoused in Uganda. Indian courts have attempted instead to narrow the discretion conferred by the provision. The judgment of the Supreme Court in Anuradha Bhasin (internet shutdown in Kashmir) recently affirmed that a valid exercise of power under Section 144 is premised on: (i) the existence of objective material facts which form the basis of the opinion formed by the Magistrate; (ii) its general invocation being confined to a specific area and issue; (iii) the existence of a demonstrably urgent situation; (iv) such measure being the least restrictive course of action; and (v) compliance with proportionality standard.

Despite these restrictions, the burden of proof continues to rest on individuals as magistrates are empowered to impose anticipatory orders prohibiting any assembly. Even where these orders are challenged before courts, the preliminary burden falls on the individual (as a petitioner) to prove that the issuance of the order violated their fundamental right. This was most evident in December 2019 where when faced with the legitimate expression of dissent against the Citizenship Amendment Act, numerous orders under Section 144 were imposed across the country without any well-founded apprehension of violence, citing inconvenience.

Conclusion

Despite its failings, the judgment in HRNU held that there is “absolutely no legal authority” to stop peaceful expression on the basis of an alleged breach of peace [p. 22] and builds on Muwanga to add to the growing jurisprudence that restrictions on the right to free speech and assembly are exceptions which the state is required to justify in every case prior to its imposition, reaffirming a commitment to a culture of justification, not authority. With this, it also provides guidance to India in the exercise of the power under Section 144. While Section 144 remains operative, one can only hope that the government of the day responds proactively to the protection of the right to free speech and assembly in scrutinizing more closely the impositions of these orders in the first place.

The Costs of Censorship: The Cable Network Rules and the Banning of AsiaNet and MediaOne

The Information & Broadcasting Ministry’s order banning the Malayalam channels Asianet and MediaOne has yet again demonstrated the weak foundations of the legal regime dealing with television content regulation. The legal authority for the ban flows from the Programme Code, passed under the Cable Network (Regulation Act). In a series of posts, Apar Gupta has extensively discussed the constitutional problems with this legislative scheme. I will summarise them briefly:

  • The Act and the Code allocate the costs of censorship to the speaker: They do so by granting the government the power to ban channels, leaving the burden (financial and otherwise) upon the channels (then) to fight it out in Court. This effectively makes censorship costless for the government, as all it has to do is pass censorship orders; contrast this with a situation in which the government would first have to prove a legal violation in court, and a judicial order was required for banning channels. This would place the initial burden upon the government (and not on the citizen), and would create a judicial safeguard before speech could be taken out of the marketplace of ideas, instead of after. Now in the wake of Puttaswamy and the advent of the proportionality standard, I would argue that it is at least arguable now that this choice of legal architecture (government bans that have to be then contested in Court by TV channels) is evidently not the “least restrictive method” of achieving the government’s goals of ensuring compliance with the law. When a less restrictive method – judicial sanction (except arguably in cases of emergencies) is available, the Act and the Code fail the test of proportionality.
  • The chilling and the conforming effect: as a related point, the legal architecture of the Programme Code creates both a chilling effect and what free speech scholar Margot Kaminski calls a “conforming effect.” That is, given that censorship is costless (and challenging it is costly), a number of bans of this kind will go unchallenged in Court. What this leads to is a “conforming effect”, where grounds set out in one banning order serve as signals to other channels to comply with them, whatever the legality might be. For example, the present orders cite apparent bias against the Delhi Police and the Rashtriya Swayamsevak Sangh (RSS) as grounds for the ban. It should be abundantly clear that this is grossly illegal; as the illegality, however, is effectively costless to the State (at best, the TV channels will fight the ban and get it overturned a few months – or years – later), it can do it again, and again – in the future. Channels know that fact just as well, and are likely to modify their behaviour to not criticise the police or the RSS, if the likely consequence is going to be an immediate 48-hour ban. This too infringes Article 19(1)(a) of the Constitution.
  • Vagueness: This is, by now, almost tedious to repeat. The provisions of the Programme Code are exceedingly vague, leaving its implementation almost entirely at the mercy of the government. Once again, this feeds into the first point about legal architecture: if, at the first instance, a Court was examining the question, then presumably – even to these vague provisions – it would apply constitutional tests (such as the incitement standard for public order). But this is not the case when it comes to the structure of the Act and the Code. The original sin in this case lies in the judgment of K.A. Abbas, where a similar legal structure – including the direct transplantation of Article 19(2) into the statute – was held to be constitutional. Perhaps, in the age of the proportionality standard, it is time to reconsider that as well.
  • In this case, specifically, the impact of the vague provisions of the Programme Code is seen particularly vividly, as the banning orders take advantage of its loose language, and mirror it with even looser language (a point discussed in the previous post). Phrases such as “biased”, “siding with a particular community”, “promoting anti-national attitudes” are so boundlessly manipulable, that they can effectively mean whatever those with brute power want them to mean. Note that this is not a case of the government abusing the law; this is a case of how a badly-drafted law enables unconstitutional use by the State. This is why both the Act and the Code ought to be struck down – it is long overdue.

Defining the Political: The Supreme Court’s FCRA Judgment

[Editor’s Note: Justice is an indivisible concept. We cannot, therefore, discuss contemporary Supreme Court judgments without also acknowledging the Court’s failure – at an institutional level – to do justice in the case involving sexual harassment allegations against a former Chief Justice. This editorial caveat will remain in place for all future posts on this blog dealing with the Supreme Court, until there is a material change in circumstances.]


The Foreign Contribution (Regulation) Act – as the name suggests – regulates the circumstances under which individuals or bodies can accept funding from foreign sources. Section 3(1)(f) of the Act prohibits any “organisation of a political nature” from accepting foreign funds. The power to specify an organisation as a “political organisation” flows from Sections 5(1) and 48; the Central Government, having regard to the “activities”, “ideology”, or “association … with the activities of any political party” can specify that an organisation is of a “political nature.” To further concretise this, the FCRA Rules of 2011 set out a number of guidelines. According to Rule 3 of the 2011 Rules, organisations with “avowed political objectives” in their MoA or bye-laws, Trade Unions promoting “political goals”, action groups with objectives of a “political nature”, organisations aiming to advance “political interests”, and organisations using “common methods of political action … in support of public causes”, can all be declared organisations of a “political nature” under Section 5 of the parent Act.

This composite scheme was challenged in Indian Social Action Forum v Union of IndiaIn a brief judgment, a two-judge bench of the Supreme Court upheld the constitutional validity of the Act and the Rules, but “read down” the last clause of Rule 3 – i.e., 3(1)(f) (“… common methods of political action … in support of public causes”) – to “active politics” or “party politics.” While the narrow reading of these excessively broad provisions is no doubt a good thing, nonetheless, in this post, I will flag three issues with the reasoning of the Court.

A. What is the “Political”

In paragraph 18 of the judgment, the Court notes that “preventing foreign contribution into the political arena is the object sought to be achieved by the Act. Prevention of foreign contributions routed through voluntary organisations which are not connected to party politics is the reason behind introduction of Section 3 (1) (f) and Section 5 of the Act.” Immediately after that, the Court goes on to note that “as the intention of the legislature is to prohibit foreign funds in active politics, an Association with avowed political objectives (i.e. to play a role in active politics or party politics) cannot be permitted access to foreign funds.”

Thus, the Court limits the scope of the use of the word “politics” across the Rules (and not just in Section 3(1)(f)) to “party politics” or “active politics.” Unfortunately, however, the Court fails entirely to define what “active politics” mean (especially as distinguished from “party politics”). What one can glean from the judgment – and especially the Court’s reference to “party politics” in its discussion of the legislative intent, and subsequently – in paragraph 21, the reference to “administration” – is that the purpose of the FCRA is to prevent foreign interference in electoral politics, so that the governance of the country is not affected by foreign interests. In this sense, “active politics” is probably best read alongside “party politics”, and – more broadly – as a prohibition upon organisations that seeks funds for electoral purposes (whether through party politics or otherwise).

If this is the meaning of “active politics”, however, then it should have been clarified. This is especially true because the word “political” is capable of boundlessly wide meaning. Indeed, as feminist thought has shown us over the years, the history of many struggles and movements is the history of attempts to shift the line between that which is “political” (and therefore subject to democratic norms, and ideas of liberty and equality), and that which is “private”. The core problem with the FCRA is the manner in which the word “political” is used throughout the statute and the Rules, without any indication of the work that it is meant to be doing. While “party politics” is at least an identifiable and specific narrowing down of the word, “active politics” has the potential to only multiply the confusion (unless, potentially, it is read in the manner suggested above).

B. Vagueness, Over-Breadth, and Abuse

This reluctance to be specific affects the Court’s judgment throughout, and leads to a serious misstep in paragraph 19, when it is addressing Rule 3(v) (“organisations … addressing political interests.” The Court holds:

We are in agreement that the words ‘political interests’ are vague and are susceptible to misuse. However, possible abuse of power is not a ground to declare a provision unconstitutional.

Unfortunately, this is a basic conceptual error that turns the doctrines of vagueness and over-breadth on their head. When a statute or a rule is attacked on the grounds of “over-breadth” or “vagueness”, the argument is not that it should be struck down because there is a “possible abuse of power.” The argument is that the language of the statute or rule is either broad enough or vague enough so as to encompass both constitutional and unconstitutional application within the terms of that language. The problem is not, therefore, the unconstitutional abuse of the law, but its unconstitutional use. As the Supreme Court of the United States noted in Grayned v Rockford, the judgment that first articulated the vagueness standard with clarity:

Vague laws offend several important values. First, because we assume that man is free to steer between lawful and unlawful conduct, we insist that laws give the person of ordinary intelligence a reasonable opportunity to know what is prohibited, so that he may act accordingly. Vague laws may trap the innocent by not providing fair warning. Second, if arbitrary and discriminatory enforcement is to be prevented, laws must provide explicit standards for those who apply them. A vague law impermissibly delegates basic policy matters to policemen, judges, and juries for resolution on an ad hoc and subjective basis, with the attendant dangers of arbitrary and discriminatory application. Third, but related, where a vague statute “abut[s] upon sensitive areas of basic First Amendment freedoms,” it “operates to inhibit the exercise of [those] freedoms.” Uncertain meanings inevitably lead citizens to “‘steer far wider of the unlawful zone’ . . . than if the boundaries of the forbidden areas were clearly marked.”

 

It should therefore be clear that once there is a judicial finding of vagueness of over-breadth, it is that finding itself that provides the reasons for unconstitutionality. The point is, as Grayned points out, that a vague or over-broad statute provides plausible legislative cover for unconstitutional State action. Or, in the words of Chintaman Rao v State of MP, a judgment that this bench does not cite: “so long as the possibility of [a statute] being applied for purposes not sanctioned by the Constitution cannot be ruled out, it must be held to be wholly void.”

C. Reading Down

A final, related point: the Court’s chosen strategy in this case (as indicated above) is to “read down” the word “politics” wherever it is found in the statute and the Rules to “party politics” or “active politics.” I have argued above that the reading down method was inapplicable in this case, as a matter of law. There is, however, a more pragmatic point that needs to be made. The strategy of “reading down” works in a situation where – after the reading down – there are mechanisms to swiftly correct the abuse of law if and when the government continues acting in the same old way, and does not follow the Court’s interpretation of the statute. As we have repeatedly seen, however – and most recently, in the case of sedition – Supreme Court judgments that “read down” legal provisions (and that, effectively, create a gap between what the text appears to mean and what the Supreme Court says it means) that are otherwise clearly unconstitutional – very quickly turn into dead letters. Without the existence of some mechanism to ensure that the “reading down” is actually effective, judgments such as these appear to be bringing the government to account, but their net impact is negligible in terms of enforcement. That is a pragmatic reality that appellate Courts should be taking into account when they fashion remedies in cases of this kind.

A Very Strange Constitution Bench

[Editorial Note: Justice is an indivisible concept. We cannot, therefore, discuss contemporary Supreme Court judgments without also acknowledging the Court’s failure – at an institutional level – to do justice in the case involving sexual harassment allegations against the Chief Justice. This editorial caveat will remain in place for all future posts on this blog dealing with the Supreme Court, until there is a material change in circumstances.]


More than three years ago – in August 2016 – I’d written about an odd Supreme Court hearing where “questions” were framed (inter alia) about whether the freedom of speech could be restricted on grounds of “constitutional compassion.” The case had come to Court when UP politician Azam Khan had labeled a gangrape as a “political controversy”, and then-Justice Dipak Misra had taken the opportunity to also ask whether political functionaries had a more restricted right to free speech than ordinary citizens.

Subsequently, Azam Khan apologised to the Supreme Court, and his apology was accepted. However, the case continued. In an order dated 5th October 2017, the matter was referred to a Constitution Bench (!) to decide. On that same day, Mr Fali Nariman and Mr Harish Salve, acting as amici in the case, framed further “questions” to be decided. Among Mr. Nariman’s questions was the following:

Whether, and if so under what circumstances (if any) would a private individual or group of private individuals (including private corporations) be required to conform to the rigor and discipline of Article 21 (in the Fundamental Rights chapter) of the Constitution – whether as “State” as broadly defined, or otherwise.

CJI Misra himself retired last year, with no further progress on this case. One would have thought, then, that this little piece of judicial buccaneering would meet a quiet and deserved end. But no: it turned out last week that the Constitution Bench of the Supreme Court – consisting of Mishra, Banerjee, Saran, Shah and Bhat JJ – would indeed sit and hear this case. And in a hearing on 24th August, events took an alarming turn, when the Attorney-General “reframed” his questions, and the Court’s order finalised them as follows: 

  1. Are the grounds specified in Article 19(2) in relation to which reasonable restrictions on the right to free speech can be imposed by law, exhaustive, or can restrictions on the right to free speech be imposed on grounds not found in Article 19(2) by invoking other fundamental rights?
  2. Can a fundamental right under Article 19 or 21 of the Constitution of India be claimed other than against the ‘State’ or its instrumentalities?
  3. Whether the State is under a duty to affirmatively protect the rights of a citizen under Article 21 of the Constitution of India even against a threat to the liberty of a citizen by the acts or omissions of another citizen or private agency?
  4. Can a statement made by a Minister, traceable to any affairs of State or for protecting the Government, be attributed vicariously to the Government itself, especially in view of the principle of Collective Responsibility?
  5. Whether a statement by a Minister, inconsistent with the rights of a citizen under Part Three of the Constitution, constitutes a violation of such constitutional rights and is actionable as ‘Constitutional Tort”?

There are a few points I want to note about this Order. While questions (4) and (5) have at least some relationship with the original case of Azam Khan, and are relatively focused, the first three questions are simply broad-ranging enquiries into general constitutional philosophy. It is entirely unclear why they have been framed in the first place, and why a Constitution Bench needs to answer them, in the absence of a specific lis that brought them to Court. Question (2), for example, is a question about horizontal rights – a vexed and fraught question that jurisdictions across the world have been grappling with. Question (3) is a question about positive obligations – another issue that Courts all over the world have spent years engaging with, and incrementally developing jurisprudence. These questions are too complex, too multi-layered, and too poly-centric to be answered in an abstract enquiry; they don’t become any simple just because a Senior Counsel has “framed” them. To take just one example: if you hold that Article 21 is indeed horizontally applicable, then how does that affect the entire swathe of Indian criminal law, whose task is precisely to protect the life and liberty of individuals from encroachment by other individuals, through legislation? Does it even affect criminal law? Should it? Legal scholars have written entire books grappling with the complex engagement between horizontal constitutional rights and existing private and public law regimes. And therefore, the correct way of addressing these issues – and the correct way for constitutional courts to behave –  is to address them in light of concrete cases that are called before it for determination, because it is only then that it can actually be seen how horizontal rights – or positive obligations – play out in the real world. Without that, this is nothing more than whistling in the dark.

But if questions (2) and (3) raise the prospect of a headless jurisprudence, question (1) is profoundly dangerous. To start with, where on earth did it come from? It wasn’t in the original four questions framed by Dipak Misra J.’s court. It is not connected to questions (4) and (5), which are the ones that are really about what speech a minister may or may not engage in, and which basically deal with the question of how and when you can attribute a minister’s speech to the State. But most importantly, this question has already been answered. It has already been answered by the Supreme Court. It has already been answered by the Supreme Court multiple times. In fact, it has been answered by a Constitution Bench of the Supreme Court. In Sakal Papers v Union of India, a five-judge bench held:

It may well be within the power of the State to place, in the interest of the general public, restrictions upon the right of a citizen to carry on business but it is not open to the State to achieve this object by directly and immediately curtailing any other freedom of that citizen guaranteed by the Constitution and which is not susceptible of abridgment on the same grounds as are set out in clause (6) of Article 19 … Freedom of speech can be restricted only in the interests of the security of the State, friendly relations with foreign State, public order, decency or morality or in relation to contempt of court, defamation or incitement to an offence. It cannot, like the freedom to carry on business, be curtailed in the interest of the general public … For, the scheme of Article 19 is to enumerate different freedoms separately and then to specify the extent of restrictions to which they may be subjected and the objects for securing which this could be done. A citizen is entitled to enjoy each and every one of the freedoms together and clause (1) does not prefer one freedom to another. That is the plain meaning of this clause. It follows from this that the State cannot make a law which directly restricts one freedom even for securing the better enjoyment of another freedom.

Sakal Papers is one of those landmark judgments in Indian free speech jurisprudence. The fundamental propositions that it articulates have never been seriously doubted in sixty years, and were accepted by the Supreme Court as recently as Shreya Singhal (2015). Most importantly, as a five-judge bench, it is binding upon the present Constitution Bench. There has been no referral to reconsider its correctness. So then, why was this question framed? Why did the five learned judges on the Bench accept it? Perhaps we’re going to find out shortly that Sakal Papers has been per incuriam these six decades, and we just didn’t know it? Who knows.

In short, therefore, a case about a Minister making a statement about a gang-rape – crass and condemnable as it was – has somehow morphed into a Constitution Bench hearing where the Supreme Court is going to opine about horizontal rights, positive obligations, and the prospect of additional limitations on free speech, all in the abstract (as the original lis has long been lost in the mists of time). And that, in a nutshell, is the problem with what PIL has become today:  it is no longer about expanding the concrete fundamental rights of the most vulnerable, but something else entirely, something that has become hostage to the fancies of individual judges and senior lawyers.

Indeed, the history of this case reveals something important about that last relationship. The history of these orders show that while the initial four questions were framed by the Court, the questions that the Constitution Bench is now going to “answer” have been framed by the two amici mentioned above. While Mr. Fali Nariman was appointed initially by the Court, Mr. Salve’s appointment came about in an order dated 29.3.2017, which records that “Mr. Harish Salve, learned senior counsel, who was present in Court, expressed his intention to assist the Court.”

Now I wonder how many of us are accorded this unique privilege of simply standing up and “expressing our intention” to assist the Court, being promptly being made an amicus, and then framing questions that are going to affect fundamental rights across the country. And this is not the first time: in the famous 1997 phone tapping case, PUCL v Union of India, while PUCL’s lawyer, Mr. Sanjay Parikh was arguing, there was another courtroom intervention. The judgment records it thus:

At this stage, Mr. Kapil Sibal & Dr. Dhawan, who are present in Court, stated that according to them the matter is important and they being responsible members of the Bar, are duty bound to assist this Court in a matter like this. We appreciate the gesture. We permit them to intervene in this matter. They need a short adjournment to assist us.

 

Mr. Sibal then went on to justify telephone interception and proposed a “non-judicial” oversight mechanism – a proposal that was accepted by the Court, and continues to haunt our surveillance jurisprudence till today. Similarly, in the notorious Devidas Tuljapurkar case, the framing that “historically respectable personalities” should be given some kind of different protection against “obscene speech” was propounded by Mr. Nariman, who had been appointed as an amicus.

None of this is the mark of an egalitarian, democratic, or even healthy legal culture, where seniority has become a kind of special-access gateway (for more on the outsize role of the amicus in PIL cases, see Anuj Bhuwania’s Courting the People). And apart from its iniquitous character, it is particularly problematic in civil rights cases, where what is at stake are fundamental liberties: these fundamental liberties should not become playthings of whatever fancy takes a Senior Counsel at a particular time. The first three “questions” that this Constitution Bench of five learned judges is now proposing to “answer” shows us the perils of such a legal culture.

Freedom of Speech at the High Courts: Contrasting Decisions from P&H and Bombay

Two decisions, delivered over the last few days, exemplify how the terrain of free speech remains a contested field in Indian constitutional law.

Burdening Legal Speech: Vishal Dadlani and the P&H High Court

Vishal Dadlani v State of Haryana came to the Punjab & Haryana High Court as a quashing petition. FIRs had been registered against Vishal Dadlani and Tehseen Poonawala under Sections 153A (promoting enmity between classes), 295A (hurting religious sentiments), and 509 (insulting the modesty of a woman) of the IPC, and 66E of the IT Act (publishing images of private body parts). These FIRs arose out an event in 2016, when the Haryana Legislative Assembly invited Tarun Sagar, a Jain monk, to deliver an address; according to the tenets of his faith, he did so in the nude. On Twitter, Dadlani criticised this strongly, focusing his ire upon the mixing of politics and religion. After getting into acrimonious arguments with Tarun Sagar’s followers, Dadlani apologised.

Meanwhile, Tehseen Poonawala criticised the prevalence of sartorial double standards, asking “why is this naked man ‘holy’ even if he walks nude in the state assembly while a woman would be termed a slut?” To drive the point home, he also posted a photograph of a “half-naked woman”, placed alongside a photograph of Tarun Sagar. From a perusal of the tweets, it appears that at least partially, Dadlani and Poonawala’s outrage was triggered by Tarun Sagar making various remarks about the role and place of women in society, such as asking wives to “accept the discipline” imposed by husbands.

Counsel for both petitioners argued that none of the offences were made out, even prima facie (counsel for Dadlani made the additional point that he had apologised directly to Tarun Sagar, and Sagar had accepted his apology; consequently, nobody from the Jain community had taken any action – the complainant at whose behest the FIR was registered was an outsider).

After extracting the submissions of both parties (including relevant precedent), the High Court held that the quashing petitions ought to be allowed, for (broadly) the following reasons: first, none of the followers of the Jain religion had initiated proceedings, or had come forward as witnesses; secondly, it had been long-accepted in Indian jurisprudence that the right to free speech included the right to express socially and culturally unpopular or unorthodox views; thirdly, the mens rea required for offences under Sections 153A and 295A had not been demonstrated; and fourthly, after two and a half years, the State had not granted sanction to prosecute, clearly demonstrating its own lack of interest in pursuing the case.

The High Court, therefore, announced that it would quash the FIRs; but in the last two pages of the judgment, it suddenly veered off into unorthodox territory. The Court noted that “justice” would also have to be done to the Jain community; it then observed:

If the contribution made by the petitioners towards poor people is compared to the contribution made by Jain Muni Tarun Sagar, it is apparent that the petitioners have played a mischief to gain publicity without having much to their credit. In recent years, the country has witnessed large scale violent protest on incitement made by using social media platform, thereby, causing extensive damage to public property. However, the preachings of Jain Muni Tarun Sagar about non-violence, sacrifices and forgiveness, has avoided repetition of such like protest. Therefore, it would be appropriate to impose the costs of 10 lacs each on the petitioner – Vishal Dadlani and the petitioner – Tehseen Poonawala, so that in future they may not mock at any head of a religious sect, just to gain publicity on social media like Twitter. (p. 38)

In other words, therefore, the High Court made the quashing of the FIRs conditional upon both petitioners paying Rs. 10 lakhs as costs.

Now, a few observations need to be made about this order. The first is that this order is incoherent. Effectively, it tells the petitioners, “you have committed no crime, but unless you pay Rs 10 lakhs, criminal proceedings against you will continue.” This cannot be. Once the High Court returned the legal finding that the offences under Sections 153A, 295A, 509 IPC and 66E were not made out even prima facie, then it followed from that finding that the FIRs would have to stand quashed. Accepting the alternative proposition would mean that if the petitioners refused to pay Rs. 10 lakhs, then the FIRs would continue and the case would go to the stage of charge and trial despite a judgment by the High Court stating that the offences had not been made out. This would be patently absurd – what, exactly, would a trial court be expected to do when faced with a situation like this?

Apart from being absurd, the High Court’s order is also illegal. Once it had found that the offences had not been made out, the logical consequence to that finding was that the petitioners had engaged in legal speech (however ill-thought, crass, or crude that speech might have turned out to be). Levying a financial penalty of Rs 10 lakhs upon that speech, then, is a classic case of judicial censorship: penalising speech in the absence of a law. As I have argued before at some length, this is, quite simply, impermissible under the Constitution. The only method through which speech can be penalised is through a State-made law (which is subject to judicial review) Under Article 19(2), as held in both Mirajkar and (more directly) in Rupa Ashok Hurra, judicial orders do not count as law.

But apart from being absurd and illegal, the High Court’s order also betrays some fundamental misunderstandings about Indian free speech jurisprudence, and the point of having a constitutional right to free speech in the first place. By noting that the purpose of its order is to deter people from mocking “at any head of a religious sect, just to gain publicity on social media like Twitter“, the Court lays down an extraordinarily broad and vague standard for the future. Recall that this comes after the Court had found that there was no mens rea to hurt religious sentiments or provoke enmity between classes; effectively, therefore, individuals who make comments that can be construed as “mocking the head of a religious sect”, and whose motives can be construed to be “just to gain publicity on social media”, can expect to be slapped with costs of Rs 10 lakhs. If this was the standard set out in a law, it would be immediately struck down as unconstitutional; indeed, in Shreya Singhal, similar phrases such as “grossly offensive” and “menacing” were struck down as being overly broad and vague. As the Supreme Court in Shreya Singhal understood, provisions such as these create a “chilling effect”: they blur the line between what is permitted and what is prohibited to such an extent, that people begin to self-censor, in order to steer far clear of that (now) invisible line.

There are two more serious mistakes in the High Court’s reasoning. The first is an implicit comparison it makes between what Tarun Sagar has done for the poor, and what the petitioners have (or have not) done. This is entirely irrelevant: the Indian Constitution does not assign value to free speech based upon the social qualities of the speaker. Setting up a hierarchy of speakers, where those who have done social work have more free speech rights than those who haven’t, is entirely inconsistent with the very concept of a right, apart from failing to understand that the Constitution protects free speech for reasons flowing from individual self-determination, democracy, and the search for truth, none of which have anything to do with the qualities of the speaker. And secondly, the High Court’s reference to “large scale violent protest” because of “incitement” on social media sanctifies the heckler’s veto: it places the burden of preventing riots on those who speak, rather than upon those who riot. That cannot be, as the Supreme Court held in Rangarajan.

It should therefore be clear that, both from the perspective of legality, and form the perspective of the deeper constitutional principles underlying the free speech guarantee, the High Court’s order is unsustainable, and ought to be promptly overturned on appeal. 

Defamation and Injunctions: The Bombay High Court

Meanwhile, in Lodha Developers Ltd v Krishnaraj Rao, the Bombay High Court considered a (typical) injunction application in a defamation case. The plaintiff was a real estate developer. Defendant No. 1 was a journalist, who had written critically about the plaintiff. Defendant Nos. 2 and 3 were purchasers who bought flats from the Plaintiff; they found the Plaintiff’s conduct and quality unsatisfactory, and criticised it online (according to the Plaintiff, it “went viral”). The plaintiff then filed a defamation suit. The defendants (as the Court noted in paragraph 11, pleaded the defence of justification (i.e., truth)). In this particular proceeding, however, the injunction was requested only against Defendant No. 1, and for five specific statements: an allegation that the plaintiff was “in connivance” with the MMRDA officials, that there was “golmaal”, that “norms” had been thrown to the winds, that “banks are part of the Lodha scam”, and that there was no occupation certificate.

Gautam Patel J. refused to grant relief for any of the five statements. His reasoning repays close study. For each of the statements, he began by carefully distinguishing whether they fell in the category of “facts”, or of “comments”. This distinction is crucial to defamation law, but is (unfortunately) far too often ignored in practice. The reason for this is that there are two separate defences in the common law of defamation: justification, and fair comment. Justification (or truth) applies to factual statements. Fair comment applies to opinion; now, contrary to what it sounds at first blush, the term “fair comment” does not mean that the comment must be “fair”, but merely that it must be founded on some factual basis that is of relevance to it. Once the factual basis is set out – and it is made clear that the comment pertains to those set of facts – then the defence applies, even though the comment itself might be hyperbolic or excessively dramatic. The reason for this is that once the reader has the facts in front of her, and has the comment, then she is best-positioned to judge whether the comment is warranted on the basis of the facts or not.

Applying this distinction, Patel J. noted that the first, second, and fourth statements were comments (and for the second comment, the Defendant promised not to repeat it without setting out his facts); and the third and fifth were factual statements. For the first and fourth statements, Patel J. held that the Defendant had set out the factual context from which he had drawn his inferences. Consequently, no injunction could be granted.

This brought the Court to the third and fifth statements, where the defence of justification (truth) was pleaded. The Court then considered the legal standard that was to apply when an injunction was sought on the basis of disputed facts. On a review of precedent (paragraph 26 onwards), the Court correctly concluded that at interim stage, an injunction could only be granted if the defences were bound to fail at trial. Consequently, as long as the Defendant produced some material in defence of her factual statements, the case would go to trial, but at the interim stage, she would not be injuncted.

This is an important observation, for many reasons. First, it adds further heft to the Delhi High Court’s detailed judgment in Tata Sons v Greenpeace, where, after a detailed survey of the common law of defamation, and in the context of Article 19(1)(a) of the Constitution, Justice Ravindra Bhat had arrived at an identical conclusion. Secondly, it comes at a particularly important time, when broad-based injunctions, granted at the ex-parte ad-interim stage, are becoming increasingly common (the Tejaswi Surya gag order, which was later set aside by the Karnataka High Court, being the latest example). By restating the law, Patel J. reminds us that at the interim stage, the law of defamation is actually tilted towards the speaker (as it should be, because the correct remedy for defamation, being a tortious offence, is compensation, not gagging). It is not for judges to take a look at the plaint, find themselves shocked at the seeming attacks upon a person’s reputation, and immediately proceed to issue gag orders; rather, it is for judges to remember that the rule is that speech is to be met with counter-speech pending trial, and a gag is the very rare exception. And thirdly, the judgment reveals the flaws in the reasoning of Shri Maheshwar Hydel Power Corporation v Chitroopa Palit, a 2004 judgment single-judge judgment of the Bombay High Court, which also concerned defamation and injunctions. In Palit, however, the Court articulated a significantly higher threshold at the interim stage, including (among other things) a requirement for the defendant to show public interest (note that “public interest” is a requirement under criminal defamation law, not civil). In the present case, Patel J. observed that even the tests under Palit had been satisfied; his own articulation of the standard, however, as we have seen, falls on the Tata Sons v Greenpeace side, and raises the hope that in due course, Palit will become an outlier judgment, as more and more cases endorse the (correct) Greenpeace view.

One final point: it was argued before the Court that the harm had been accentuated because the defamatory statements had gone viral on social media. This form of argument – that holds that legal standards protecting speech should be diluted based on the medium – is a familiar one; recall that it was made in Shreya Singhal as well, and the judgment itself is unclear on what stand it takes. Patel J., however, was unequivocal: the medium made no difference to the legal and constitutional standards at issue. Indeed, he made the (additional) important observation that the plurality of voices in the online world made the requirement of tolerating opposed views more urgent, rather than less.

Conclusion

The contrasting judgments – and approaches – taken by the Punjab & Haryana and the Bombay High Courts remind us that free speech adjudication remains highly judge-centric. A part of the reason for this seems to be that the separation between speech that we perceive to be irritating, value-less, mischievous – in a word, rubbish – and speech that is illegal, is still not embedded firmly enough in our jurisprudence. This is not necessarily a criticism: notwithstanding (the quotation attributed to) Voltaire, it is but human to allow one’s contempt for a speaker, or for what they are saying, to get in the way of a dispassionate constitutional analysis. That is what seems to have happened in the P&H case, with the pointed references to the petitioners’ (lack of) social work, and their desire for fifteen minutes of fame on twitter. But the future of Indian free speech jurisprudence depends upon judges being able to make that distinction; hopefully, they will have that opportunity soon enough.

Making the Path by Walking: The Supreme Court’s Film Censorship Judgment

In an interesting judgment handed down this Thursday, a two-judge bench of the Supreme Court held that the government of West Bengal was required to compensate a film-maker for trying to “shadow-ban” his film. The facts of Indibility Creative v Govt of West Bengal were straightforward: a film called Bhobishyoter Bhoot had been cleared by the Film Certification authorities, and had started to run in cinemas. However four days before its scheduled release, one of the directors received a call and a letter from the state’s intelligence unit, asking for a special screening, on the ground that the film might hurt public sentiments and threaten public order. The director refused. However, very soon after the release of the film, it was inexplicably withdrawn from most cinemas, and tickets were refunded. When the directors made enquiries, they were told that this was being done on the instructions of “higher authorities.”

After the Supreme Court intervened with some strongly-worded interim orders, a modicum of normalcy was restored. The Court nonetheless reserved orders, and delivered a judgment on merits shortly thereafter.

Chandrachud J.’s judgment involves a reiteration and endorsement of settled legal principles that are, regrettably, too often honoured only in the breach: that there cannot be a two-track certification process, where the State’ authorities are running parallel censorship proceedings alongside the Certification Board(s); that in cases where there is an apprehension of public disorder, it is incumbent upon the State to provide the necessary police protection, rather than upon the speaker/artist/writer to withdraw (and thereby facilitating the heckler’s veto); and that the restrictions upon free speech cannot traverse beyond the strict confines of Article 19(2). In addition, however, the judgment contains a few elements worthy of remark.

The first is that it takes seriously the ways in which State authorities can informally exert pressure and effectively choke off the meaningful exercise of fundamental rights. In this case, for example, the State government insisted that it had not taken recourse to any statutory provision that formally banned the film. If that was true, however, then, as the Court observed:

… there has to be some explanation forthcoming before the Court why the film was simultaneously removed from the theatres, at one stroke, shortly after release. The apprehension of the petitioners that this was an action which followed on the letter dated 11 February 2019 of the Joint Commissioner of Police is not unfounded. The letter addressed by INOX to the producer specifically mentions that they were directed by the authorities to discontinue the screening in the ‘interest of the guests’. We have no manner of doubt that this was a clear abuse of public power. (paragraph 17)

Effectively, therefore, the Court drew an adverse inference against the State, based upon a set of fairly unambiguous factual circumstances. Following upon the Madras High Court’s recent judgment involving the “informal” settlement that Perumal Murugan was forced into, this signals a clear judicial intent to take seriously State action that would otherwise pass under the radar due to its non-legal/extra-constitutional/behind-the-scenes character.

This is closely linked to a second important point – the Court’s clear articulation of the positive component of Article 19(1)(a). As Chandrachud J. noted:

But, apart from imposing ‘negative’ restraints on the state these freedoms impose a positive mandate as well. In its capacity as a public authority enforcing the rule of law, the state must ensure that conditions in which these freedoms flourish are maintained. (paragraph 18)

This is, of course, nothing more than applying a coherent constitutional framework to the basic insight – expressed first in S. Rangarajan and then beyond – that if the threats to free speech come from private parties, then it is the obligation of the State to ensure that conditions of law and order are maintained so that the expressive activity in question can go ahead. Rangarajan’s insight, of course, arose out of concrete circumstances where the State threw up its hands before the Court and argued it was not in a position to guarantee the maintenance of law and order if a particular film was screened – an argument that was given short shrift by the Court. Indibility Creative – the present judgment – has now taken that insight further, and generalised it into a principle.

The third element that rounds things off is the relief: the Court held that:

As a consequence of the pulling off of the film from the theatres where it was screened on 16 February 2019, the petitioners have suffered a violation of their fundamental right to free speech and expression and of their right to pursue a lawful business. This has been occasioned by the acts of commission and, in any event, of omission on the part of the state in failing to affirm, fulfill and respect the fundamental freedoms of the petitioners. We are clearly of the view that a remedy in public law for the grant of remedial compensation is required in the present case. We order and direct the respondents to pay to the petitioners compensation which we quantify at Rs 20 lakhs within a period of one month from the date of the present judgment. (paragraph 20)

This is particularly important, because in far too many cases, after finding a breach of fundamental rights, the Court has left matters as they are. The loss, however – especially in cases of this kind – is clear and quantifiable, and a refusal to restore status quo only ensures that, despite the finding of a violation by the court, the chilling effect of arbitrary State action continues. On the other hand, a clear judgment on compensation sets the tone for an effective enforcement of fundamental rights; and it is only within a climate in which enforcement is guaranteed, can those rights truly flourish.

Thus, a combination of these three features – judicial cognisance of informal, “shadow-banning”, an articulation of the positive obligations upon the State to ensure that free expression is guaranteed notwithstanding threats from private parties, and remedial compensation should the State fail to discharge its obligations – ensure that this judgment can become the basis of a strong, rights-protective jurisprudence. Whether that happens, of course, depends on how future courts – especially the lower courts – implement it.

A final point of interest. In a paragraph dealing with the importance of satire to any system of free expression, Chandrachud J. observed that:

Satire is a literary genre where “topical issues” are “held up to scorn by means of ridicule or irony.”It is one of the most effective art forms revealing the absurdities, hypocrisies and contradictions in so much of life. It has the unique ability to quickly and clearly make a point and facilitate understanding in ways that other forms of communication and expression often do not. However, we cannot ignore that like all forms of speech and expression, satirical expression maybe restricted in accordance with the restrictions envisaged under Article 19(2) of the Constitution. For example, when satire targets society’s marginalized, it can have the power to confirm and strengthen people’s prejudices against the group in question, which only marginalizes and disenfranchises them more. (paragraph 13)

The last lines are particularly important, because they point towards the Canadian/South African constitutional model of understanding hate speech: recall that in Canada and South Africa, “hate speech” is asymmetrical, and is understood as expression that stigmatises or dehumanises a vulnerable class of people, in a manner that feeds into a wider climate of marginalisation and discrimination (for example, holocaust denial as a form of anti-semitic speech, or – closer home – caste slurs such as “chamar”). The regulation of hate speech, thus, is inextricably bound up with the constitutional value of equality (expressed in that old injunction to satirists and critics, “punch up, not down“).

In India, that model is yet to catch hold; while the Court hinted at that in 2014, for the most part, the dominant understanding remains that of hurt sentiments and wounded feelings – an approach that, for obvious reasons, is constitutionally unsustainable. And so, while the observations in today’s judgment are fleeting, they do constitute an invitation to a future bench, dealing with an appropriate case, to make them a part of the law.

We can hope that some day, that invitation will be taken up.

 

Guest Post: On Free Speech and Jurisdictional Issues in Online Defamation Cases

(This is a guest post by Raghav Kohli.)

An overwhelming cavalcade of technological innovations since the last century has redefined the relationship between law and technology; the Indian Judiciary, quite helplessly, has been playing catch-up ever since. Recently, in one of many such instances that demonstrate the awkwardness of Indian judges in engaging with technology, the Supreme Court in M/s Future Gaming and Hotel Services Pvt. Ltd v. Malayala Manorama & Ors passed an order confirming the regressive approach of Indian Courts towards jurisdictional issues arising out of the online dissemination of free speech.

The facts, briefly, were as follows: The Respondents, a Malayalam Vernacular Daily Newspaper, published a news article on 22-04-2015 that was also made available online on their website. A complaint was lodged by the Petitioner, before the Chief Judicial Magistrate, East and North, at Gangtok (hereafter CJM), inter alia, on grounds that the news so published had defamed the Petitioner Company. The CJM issued process of summons against the Respondents, who subsequently filed a Petition under Article 482 of the Code of Criminal Procedure (hereafter CrPC) before the High Court of Sikkim at Gangtok for quashing the complaint and setting aside the summon order. The High Court quashed the complaint for want of territorial jurisdiction of Courts in Sikkim as the Complainant had “failed to point to any person who has in fact read the online version or downloaded the same to make the offence under Section 499 of the IPC complete and thereby extend jurisdiction to the Courts in Sikkim” (paragraph 9).

Consequently, the Petitioners filed a Special Leave Petition before the Supreme Court, which set aside the decision of the Sikkim High Court. It was held that the complaint could not be quashed at this stage since the list of witnesses contained residents of Sikkim, who proposed to give evidence to the effect that the alleged offence had taken place in Sikkim. The division bench of AK Goel, J. and UU Lalit, J. thus impliedly confirmed that in cases of online defamation, courts could assume jurisdiction in any place where the impugned material is accessed.

The Law

Before addressing the question of why such an approach is problematic, it is of pertinence to analyse the rules governing the territorial jurisdiction of courts in civil and criminal cases. Among other provisions to determine jurisdiction in Chapter XIII of the CrPC, Section 179 provides that “When an act is an offence by reason of anything which has been done and of a consequence which has ensued, the offence may be inquired into or tried by a Court within whose local jurisdiction such thing has been done or such consequence has ensued.” Similarly, Section 19 of the Code of Civil Procedure (hereafter CPC) provides that “where a suit is for compensation for wrong done to the person or to movable property, if the wrong was done within the local limits of the jurisdiction of one Court and the defendant resides, or carries on business, or personally works for gain, within the local limits of the jurisdiction of another Court, the suit may be instituted at the option of the plaintiff in either of the said Court”.

Traditionally, courts have interpreted Section 19 of the CPC and Section 179 of the CrPC expansively in offline defamation cases to extend territorial jurisdiction to not only Courts at the place where the defamatory statement is first made and published (in the sense of the statement being communicated), but also where it is subsequently published, circulated and read. For example, in Dr. Subramaniam Swamy vs Prabhakar S. Pai, the assailed statement was made at Chandigarh at a press conference, but also later published by the Indian Express in Bombay. The Bombay High Court held that “the consequence of the statement made at Chandigarh has been completed at Bombay by circulation of the said newspapers, and, therefore, the offence of defamation is complete in the City of Bombay. In view of the provisions of S. 179 of the Criminal P.C. both the Courts, at Chandigarh and at Bombay, will have jurisdiction to entertain a complaint under S. 500 of the Indian Penal Code.” Similarly, the Karnataka High Court in P.Lankesh and Another v. H. Shivappa & Anr., held that “It cannot be said that the act of publication comes to an end as soon as one issue of the newspaper is released at one place … If the defamatory imputation is made available to public at several places then the offence is committed at each such place. Though the first offence may be committed at the place where it is printed and first published, it gets repeated wherever the newspaper is circulated at other places.”

The Law in the Online World

The issue arises when Courts apply the same standard for determining jurisdiction in cases of online dissemination of speech; and High Courts across India have done that unanimously. In fact, the Delhi High Court as recently as in the 2016 case of Frankfinn Aviation Services Pvt. Ltd v. Tara Kerkar & Ors held that by putting defamatory material on the internet, “territorial jurisdiction does not remain confined to the place of actual defamation … (and) jurisdiction would be at both places i.e. the place where the actual defamation takes place and the place where such defamatory material is transmitted through website”. Even the Sikkim High Court in the Malayala Manorama case would have found that Courts in Sikkim are competent to exercise territorial jurisdiction if only the Petitioner had produce a person who had read (or downloaded?) the online version in Sikkim.

However, courts have conveniently chosen not to engage with the detrimental consequences of adopting such an expansive approach on free speech. By allowing for the assumption of jurisdiction by any court situated at a place where the website may be accessed, courts have failed to appreciate the inherent distinctions between traditional media and new digital media, and effectively permitted individuals to “create” territorial jurisdiction in online defamation cases as per their convenience across the country (at least). Consequently, this provision has been systematically used to harass journalists, authors, and other individuals, who are forced to travel to remote locations at great personal expense. In fact, Human Rights Watch in 2016 reported several such instances of harassment, with the Tamil Nadu government, for example, having reportedly filed nearly 200 cases of criminal defamation between 2011 and 2016.

The hardship faced by an accused is further augmented as there is no cap on the number of cases that may be filed against him or her under Section 199, CrPC, which lays down the procedure for prosecution. It is also settled law that the exceptions to defamation are only considered after the trial commences in light of Section 204, CrPC, read with Section 105 of the Indian Evidence Act (reiterated in paragraph 198 of Subramanium Swamy v. Union of India). Thus, the accused has minimal safeguards at the time of issuance of process against unscrupulous allegations. Although Section 202, CrPC mandates that the Magistrate postpone the issue of process for deciding whether or not there are sufficient grounds for proceeding in cases where the accused is residing at a place beyond his/her jurisdiction, it often does not afford adequate protection to the accused against harassment, making the process a punishment in itself. This is also applicable to the rare occasions when the accused may be allowed to appear via video conferencing (as was done in the Malayala Manorama case), which engenders its own set of issues.

The Chilling Effect

The cumulative effect of these draconian procedures is a “chilling effect” on speech, which prompts people to engage in self-protective censorship in fear of penalisation. Although this concept has formed an essential part of First Amendment phraseology and jurisprudence in the United States since the 1950s, it was adopted into our free speech jurisprudence much later (starting with the Delhi High Court). The Indian Supreme Court has now recognised this concept in several cases, such as R. Rajagopal v. State of T.N, where the Court modified the common law of civil defamation and noted the chilling effect caused by a no-fault liability standard (paragraph 19). Similarly, in S. Khushboo v. Kanniammal, the Court observed that the law “should not be used in a manner that has chilling effects on the freedom of speech and expression” (paragraph 29). Most famously, in the widely celebrated judgment of Shreya Singhal v. Union of India, the Supreme Court invoked the principles of “vagueness” and “overbreadth” in addition to the chilling effect to strike down Section 66A of the IT Act in 2015 (paragraph 90).

Interestingly, in the case of Subramanium Swamy v. Union of India, in which the Supreme Court upheld the constitutionality of criminal defamation, a similar argument was made by some of the petitioners (Read paragraph 20 of Mr. Arvind Datar’s submission here) against the procedure governing the prosecution of defamation under the the CrPC. It was argued that such rules amounted to procedural unreasonableness and imposed a chilling effect on speech, and were thus unconstitutional. The Court’s response to this submission is a classic example of how judges often conveniently disregard engaging with the submission made, and dismiss it by merely reiterating the settled law.

After noting the abuse of procedural laws and summarising the law on territorial jurisdiction contained in Sections 176-179 and Section 186 of the CrPC, the Court observed: “Thus, CrPC governs the territorial jurisdiction and needless to say, if there is abuse of the said jurisdiction, the person grieved by the issue of summons can take appropriate steps in accordance with law. But that cannot be a reason for declaring the provision unconstitutional” (paragraph 195). The Court failed to even acknowledge that the chilling effect argument made was premised upon the inadequacy of the so-called “appropriate steps in accordance with the law.”

Similarly, on the argument of considering the exceptions under Section 499 at the time of summoning the accused, the Court summarised the settled position of law that those who plead an exception must prove it, and observed: “Therefore, the argument that if the said Exception should be taken into consideration at the time of the issuing summons it would be contrary to established criminal jurisprudence and, therefore, the stand that it cannot be taken into consideration makes the provision unreasonable, is absolutely an unsustainable one and in a way, a mercurial one. And we unhesitatingly repel the same” (paragraph 198). Apparently, using “therefore” multiple times in a sentence sufficed as a justification.

Lessons from Abroad 

In fact, it was in response to similar issues arising out of the online dissemination of speech such as forum shopping and stifling of free speech that foreign jurisdictions such as the United States evolved restrictive tests to determine “personal jurisdiction” (or, the court’s jurisdiction over the parties in a suit) in online defamation cases. The Due Process Clause of the Fourteenth Amendment permits a court to exercise personal jurisdiction (even where a long arm statute exists) “over a foreign defendant when (1) that defendant has purposefully availed himself of the benefits and protections of the forum state by establishing `minimum contacts’ with the forum state; and (2) the exercise of jurisdiction over that defendant does not offend `traditional notions of fair play and substantial justice.’ Sufficient minimum contacts will give rise to either specific or general jurisdiction. General jurisdiction exists when a defendant’s contacts with the forum state are unrelated to the cause of action but are `continuous and systematic.’ Specific jurisdiction arises when the defendant’s contacts with the forum ‘arise from, or are directly related to, the cause of action.’” (Revell v. Lidov at paragraph 20) The mere accessibility of data hosted on a website in the forum state has consistently been held to be insufficient by both standards for the forum court to assume jurisdiction. The court, therefore, determines jurisdiction based on different criteria such as the active or passive nature of the website (as per the “Zippo Sliding Scale”, for example, established in the case of Zippo Manufacturing Co v. Zippo Dot Com Inc.), or, by applying the “effects test” (established in Calder v. Jones, to adjudge whether the effects caused by an defamatory article in the forum state were specifically directed and intended towards it as to confer jurisdiction upon it).

The 2002 decision of the United States Court of Appeals, Fifth Circuit, in Revell v. Lidov is a good illustration of the application of these concepts. The brief facts are as follows: Revell (a resident of Texas) sued Lidov (a resident of Massachusetts) and Columbia University (whose principal office were in New York City) in the Northern District of Texas for defamation arising out of Lidov’s authorship of an article that he posted on an internet bulletin board hosted by Columbia. The district court dismissed Revell’s claims for lack of personal jurisdiction over both Lidov and Columbia as it found the website to be “Zippo-passive”. When this decision was appealed before the Circuit Court, the issue to be determined was whether the operation of an internet site supported the minimum contacts necessary for the exercise of personal jurisdiction. It was held that owing to both the low level of interactivity of the website, and inapplicability of the “effects” test, specific personal jurisdiction could not be established in Texas. It was observed that “the post to the bulletin board here was presumably directed at the entire world, or perhaps just concerned U.S. citizens. But certainly it was not directed specifically at Texas… As these cases aptly demonstrate, one cannot purposefully avail oneself of ‘some forum someplace’; rather, as the Supreme Court has stated, due process requires that ‘the defendant’s conduct and connection with the forum State are such that he should reasonably anticipate being haled into court there.’” Similarly, it was held that general personal jurisdiction could not be established as the “maintenance of a website is a continuous presence everywhere in the world” and does not amount to a “substantial” contact.

It is also interesting to note that while Indian courts have continued to apply traditional jurisdictional rules in online defamation cases, they have increasingly turned to the progressive US standards in trademark infringement disputes. For example, in the landmark case of Banyan Tree Holding (P) Limited v. A. Murali Krishna Reddy & Anr, the Court held that it had jurisdiction over the defendant but it did not get attracted merely on the basis of interactivity of the website which is accessible in the forum state, but on the basis that viewers in the forum state were specifically targeted by the Defendant Website. In an extensive analysis of tests adopted by different jurisdictions to assume territorial jurisdiction, Justice S. Muralidhar went to the extent of noting that “While courts have more readily applied the ‘effects’ test in defamation cases [see Remick v. Manfredy, 238 F.3d 248 (2001); Noonan v. Winston Comp., 135 F.3d 85, 91 (1998)]; Revell v. Lidov, 317 F.3d 467 (5th Cir. 2002) there have been problems in its application to trademark infringement cases” (paragraph 25). This progressive approach to determine jurisdiction has been reiterated in several High Court decisions such as the 2017 decision of the Delhi High Court in Federal Express Corporation v Fedex Securities Ltd. & Ors.

Justice Brennan in the US Supreme Court decision of NAACP v. Button famously remarked that freedom of speech needs “breathing space to survive”. Adopting a regressive approach towards jurisdictional issues arising out of online speech inhibits exactly that. It is hoped that the Indian Supreme Court on a suitable occasion in the future will deliberate on these arguments, and step up to protect this freedom from being choked.