Internet Censorship in India: Peeking Under the Hood

By: Devdutta Mukhopadhyay, Internet Freedom Foundation

Following the earlier contributions to this blog series, which provided an overview of the six GNI-Internews fellows’ research projects, in this collection, fellows document their work over the last few months.

This post was originally published on Medium

For policy researchers interested in studying Internet access and usage, India represents a paradox where price barriers are low but government ordered Internet restrictions are rampant. In 2019, India had the cheapest mobile Internet prices anywhere in the world but it was also responsible for 121 out of 213 documented Internet shutdowns imposed globally. While Internet shutdowns are the most draconian tool used to control the flow of information, Indian Internet users have also experienced an increase in blocking of specific websites and mobile apps. On 29 June 2020, in what was termed by many government supporters as a ‘digital surgical strike,’ India banned 59 Chinese-origin mobile apps including Tik Tok and We Chat. This overnight ban was the first of its kind in the world and it refocused attention on India’s opaque and unaccountable website blocking process.

In our policy brief, we have tried to go beyond the headlines to understand the on-the-ground realities of Internet censorship in India by focusing on legal orders issued by government authorities to impose these restrictions. This methodological choice is motivated by lack of existing literature which analyze the legal orders through which Internet censorship is implemented. These legal orders and their content occupy a central role in litigation challenging Internet shutdowns or website blocking, and they also determine the leeway available to information and communication technology (ICT) companies to narrowly interpret the scope of restrictions. Most importantly, these legal orders provide insight into the material available to government officials at the time of censorship being authorized. They must be capable of surviving legal scrutiny on their own without further explanations or justifications being required later in time. As the Indian Supreme Court has famously noted, “Orders are not like old wine becoming better as they grow older.”

We have structured our analysis of the legal orders authorizing Internet shutdowns and website blocking in the form of case studies. Our first set of three case studies covered the full spectrum of Internet shutdowns ranging from a complete communication shutdown in the border state of Jammu & Kashmir, where even landlines were cut off, to narrower social media bans in the eastern state of Bihar. Even within individual states, we found that the severity and frequency of Internet restrictions can vary significantly. For instance, while it is commonly believed that Jammu & Kashmir is currently experiencing an Internet slowdown, we found that in districts like Pulwama, mobile Internet services were completely suspended for at least 574 hours or 24 days this year.

In our second set of case studies, we focused on two specific instances of blocking of a satirical website and an environmental campaign website to highlight how procedural safeguards like notice, pre-decisional hearing and issuance of a written reasoned order are

often flouted by government authorities to engage in secretive and arbitrary website blocking. Critically, these case studies revealed that website blocking is being ordered against a wide range of websites including popular mainstream websites and it is being implemented by different private entities in the ICT ecosystem including Internet service providers (ISPs) and domain service providers.

We decided to study Internet shutdowns and website blocking together because they fall on different ends of the same spectrum of Internet restrictions. Further, incidents like the social media bans ordered in Bihar under the Temporary Telecom Suspension Rules 2017, which were meant to authorize conventional Internet shutdowns, clearly indicated that these problems cannot be understood in silos. Finally, Internet shutdowns and website blocking both raise similar concerns about transparency, accountability, vagueness, availability of less restrictive alternatives, and exercise of these powers by unauthorized officials.

In the final section of the policy brief, we have drawn on the case studies to identify useful lessons for different stakeholders and developed recommendations, which can be implemented to mitigate the harms caused by Internet censorship in India.

RECOMMENDATIONS

For Lawyers:

(i) Collaboration: Lawyers commonly rely on legal authorities like legislations and case law to prove their client’s case. However, this may not be enough for litigation related to Internet censorship, which often involves complex technical issues that may be beyond the understanding of many judges. For instance, if a court is expected to determine if 2G speed constitutes effective Internet access in 2020, technical evidence must be produced which clearly demonstrates the impact of 2G v. 4G on web performance. Similarly, to explain the importance of the Internet as essential infrastructure that facilitates not merely speech, but also access to healthcare, education, trade and business, it is vital to produce testimonies of actual people affected by Internet restrictions and statistical evidence quantifying the losses suffered by them. For this purpose, lawyers working on cases challenging Internet restrictions must collaborate with technologists, economists, journalists, and local activists to ensure that the harm caused by Internet restrictions is undeniably proven before a court. This is particularly relevant since government authorities routinely deny any adverse impact of Internet restrictions, and unless concrete facts and figures are produced to challenge this narrative, courts may accept the government’s claims at face value.

(ii) Incrementalism and Iteration: Litigation challenging Internet restrictions may not provide immediate substantive relief since court cases can drag on for a long time and judges may be reluctant to intervene if national security interests are invoked by the government.However, even in such circumstances, incremental improvement in the status quo is possible and the threat of litigation can also nudge government authorities to at least comply with the letter of the law, if not its spirit. This kind of incrementalism, when combined with strategically planned future interventions, can provide meaningful relief to people affected by Internet shutdowns and website blocking. For instance, while much remains to be done, 4G mobile Internet was finally restored in at least two out of 20 districts of Jammu & Kashmir on 16 August 2020 after three successive rounds of litigation before the Supreme Court, which first started with the complete communication shutdown on 05 August 2019.

For Civil Society Organizations:

(i) Measurement: As we found through the case studies, it can be difficult for people to determine if service disruptions have been ordered by the government or if they are caused by technical errors. This problem is particularly acute in the context of website blocking, which is inconsistently implemented and recent research suggests that only 27.64% of blocked websites are uniformly blocked by all six major Indian ISPs. In these circumstances, the development and use of measurement tools like the Open Observatory of Network Testing (OONI) probes are necessary to get an accurate and complete picture of Internet censorship in India. Fortunately, efforts are already being made in this direction, and researchers have used the OONI probe to test and compare website blocking in different states of India and to develop their own measurement tools.

(ii) Government Engagement: Our case studies revealed that in many instances, Internet restrictions were ordered by government officials in response to credible threats of large-scale violence along the lines of religion or caste. India has witnessed several devastating communal riots since its independence in 1947, leaving thousands dead or displaced. Therefore, the concerns of local government officials who are expected to control violent and unruly mobs cannot be dismissed lightly. It is necessary for civil society organizations who are familiar with empirical research on effectiveness of Internet shutdowns to engage with government officials to explain how Internet restrictions can actually incentivize violent protests which require less communication and coordination than peaceful demonstrations. It is also important to work with government officials to highlight how technology can be used to diffuse tensions by debunking rumours and providing accurate information. Finally, there is a need for more empirical research about both positive and negative use of the Internet during communal riots in different parts of India since the Internet also plays a role in the documentation of atrocities and fact checking, which cannot be ignored.

For Information and Communication Technology (ICT) Companies

(i) Mitigation: Our conversation with an Internet service provider revealed that due to onerous license conditions and harsh penalties, companies do not have much scope to push back against government requests. The interviewee stated that even if what was being sought by the government was practically infeasible, ISPs would still confirm compliance because as they vividly put it: “The license conditions make the Department of Telecommunications a God-like figure for ISPs.” The companies also fear that they may not receive protection from the judicial branch if they challenge the government’s actions because recently, the Indian Supreme Court has come down very severely on the telecom industry by interpreting existing legal provisions in a manner that threatens these companies by requiring them to pay huge sums of money within a short period of time.

Even in democratic countries like India, certain regions may be unusually difficult to operate in due to geo-political reasons. For instance, in Jammu & Kashmir, telecom service providers have to comply with additional customer verification requirements for pre-paid subscribers and unless these requirements are fulfilled, pre-paid subscribers cannot be allowed even 2G mobile Internet access. While operating in such regions, it is very important for companies to be sensitive to hardships being experienced by the local population and to provide them appropriate concessions. For instance, when voice call services were restored in Jammu & Kashmir in October 2019, Airtel, which has the largest subscriber base in the region, agreed to waive off post-paid rental charges for the period during which these services were suspended by the government.

Despite the limitations associated with operating in a hostile legal regime, there are still some ways in which ICT companies can mitigate the adverse impact of Internet restrictions. As we saw in the case studies, government officials want to narrow the scope of Internet shutdowns but they may not have the technical knowledge and ability to do so. For instance, in Rajasthan, the Internet shutdown orders issued to prevent cheating in police constable recruitment exams only required suspension of Internet services in and around examination centres. Similarly, while imposing social media bans, the orders issued by the Bihar government clarified that only new uploads had to be blocked on YouTube, and there was no prohibition on people using the Internet to watch existing YouTube videos. In such cases, good-faith engagement may be possible between ICT companies and government officials to develop solutions to ensure the scope of Internet restrictions is narrowly limited to what is actually required.

(ii) Circumvention: ICT companies which develop circumvention tools such as virtual private networks (VPNs) should consider the needs of users in countries where threats exist outside a rule of law framework. For instance, bans on VPNs in Jammu & Kashmir have been enforced through physical violence and people have been assaulted by police officials if VPN apps were found on their phone during random searches at highway checkpoints and other public spaces. Therefore, while developing and designing their products, companies should consider the risks faced by users living in regions where human rights violations are common and offer discreetly named or designed alternatives that may stand a better chance at avoiding detection during such random searches.

(iii) Learning: Indian ICT companies would benefit from learning about international best practices adopted by their foreign counter-parts. Here, it may be instructive for them to refer to the GNI Implementation Guidelines which provide practical and detailed guidance on how companies can fulfil their human rights obligations. The GNI Company Assessment Reports which include case studies describing how different companies have handled government demands could also serve as a useful learning resource. In fact, based on these assessment reports, David Sullivan has distilled five specific recommendations for companies facing government ordered network disruptions which include seeking clarity about legal obligations, documenting and escalating all demands, narrowing the scope of disruptions, increasing transparency and collaborating with other stakeholders.

If these recommendations are put into practice by Indian ICT companies, it would significantly improve the status quo. In particular, companies should at least be more transparent and inform users if they have been subjected to government ordered censorship. As we previously noted, many Indians are not be able to determine why a website is inaccessible or why Internet services seem slow. Here, ICT companies could step in to provide their users information about which government authority has directed service disruption, under which legal provision, and for what period of time. Having this kind of information about government ordered censorship will also make it considerably easier for aggrieved individuals to promptly challenge such measures before courts, especially if the order is issued under an incorrect legal provision or by an unauthorized government official.

In conclusion, addressing Internet shutdowns and website blocking in India will require cooperation and collaboration between different stakeholders who all have a role to play at different stages of the cycle of Internet censorship. Transparency by ICT companies will facilitate strategic litigation by lawyers and research by civil society groups. The outcome of such litigation and research must then be communicated to government officials to ensure compliance with judicial directions and evidence-based decision-making respectively.