Primer on internet shutdowns and the law
November 2016
Developed by Access Now for Special Rapporteur on the promotion and protection for the right to freedom of opinion and expression David Kaye’s “Study on Telecommunications and Internet Access Sector”
Table of Contents
Primer on internet shutdowns and the law 1
Introduction 1
How common are internet shutdowns, and how do they occur? 2
Are shutdowns ever justified under international human rights law? 3
Why and when do government officials order internet shutdowns? 5
Justifications 5
National Security 6
Elections and political events 6
Protests and demonstrations 7
School Exams 8
Legal gaps 9
When there are outdated laws and overbroad definitions 9
When local laws and processes lack transparency 10
When international standards and legal instruments allow them 10
Focus on India 11
What are some impacts of internet showdowns? 12
Shutdowns threaten the right to life, and often precede atrocities 12
Shutdowns do lasting damage to democracy and political participation 13
Shutdowns slow economic development and the digital economy 13
Who is speaking out against shutdowns? 14
International institutions and experts 15
Private companies, business associations, and civil society organizations 16
#KeepItOn Coalition 16
Legislators and judges 16
Conclusion 17
Introduction
Intentional shutdowns of communications networks, applications, and services are one of the most pressing and devastating affronts to the freedoms of expression and opinion today.
At RightsCon Silicon Valley 2016, Access Now convened an international coalition to define this rapidly growing threat, producing this definition: An intentional disruption of internet or electronic communications, rendering them inaccessible or effectively unusable, for a specific population or within a location, often to exert control over the flow of information. In our interpretation, internet shutdowns occur when governments order companies to terminate or degrade public access to digital communications tools — like Twitter, SMS, or Facebook.
Calculating the broad and grave impacts that shutdowns inflict on human rights remains beyond the capacity of most observers. They include impacts to the right to life, and to participation in civil and political, social, and cultural life, as well as damage to economic development, to trust in our democratic institutions, and even to individual psychology and physical well-being.
However, we submit with certainty that as the internet and digital communications tools extend deeper into our daily lives and become more central to our societal systems, decisions regarding infrastructure, network equipment, and telecommunications service providers will have even more power to impact the realization of human rights in the digital age.
Despite the very public impact of internet shutdowns, how and why they are carried out remains shrouded in secrecy. We know little about the reasons officials order them, the procedures that telecommunications companies use to execute them, and exactly how decisions are made to start or end a restriction. In the majority of cases, government officials do not publicly disclose the legal authority that underpins their shutdown orders. If they do exist in written form, the orders rarely see the light of day.
In the absence of either government or corporate transparency, other than a few exceptional disclosures, it is difficult simply to document the sheer scope and scale of the disruptions that are currently occurring. We are left to scrutinize brief accounts by the media or terse press releases by communications ministries, executive officials, or law enforcement chiefs, to attempt to understand why human rights are being broadly restricted, and how long the blackouts will last.
This primer contains information Access Now has pieced together with partners in the stakeholder groups involved in shutdowns, representing knowledge drawn from a seven-year period of tracking the issue, starting when our organization was born during Iran’s Green Revolution and the ensuing government crackdown on the freedoms of opinion and expression online. We invite the Special Rapporteur and other stakeholders to continue researching and advocating for transparency, accountability, and putting an end to internet shutdowns.
How common are internet shutdowns, and how do they occur?
Access Now has confirmed at least 50 incidents in 2016, as of November 1, that fit our definition of an internet shutdown.1 They occurred in more than 25 countries on four continents this year, including Europe.
This far exceeds our tallies for previous years. However, we believe our estimate is conservative, and many more incidents go unreported. The way we count has a big impact on that number as well: we count ongoing shutdowns, like Gabon’s nightly “internet curfew” in September, as a single event, rather than 14 or more individual shutdowns, for example.
Fortunately, we see increasing global awareness of the problem of shutdowns among stakeholders like journalists, internet and messaging app companies, foreign ministries, technologists, and network measurement firms. However, this increased attention has not yet resulted in an appreciable decrease in the rate of shutdowns, nor in meaningful dialogue between governments and stakeholders in the countries where they are most often deployed.
There are several ways to shut down the internet or block communications tools. One way is to interfere with the resolution of URLs to Internet Protocol (IP) addresses, so that when you type in a web address, like "https://freedex.org," your internet service provider redirects you elsewhere or your request never reaches the intended site. Another way is when an internet service provider interferes with routing tables and border gateway protocols, so that a single site, like Facebook.com, is unreachable, or an entire country no longer appears in global address books. There are many more ways, and governments are using increasingly sophisticated methods to disrupt communications -- often with the assistance of private companies.2
Are shutdowns ever justified under international human rights law?
Restrictions on freedom of expression must be provided by law, pursuant to a legitimate aim, and necessary and proportionate to achieve that aim.
We submit that:
The standard three-part test legal test for restrictions on the freedom of expression applies to internet shutdowns.
No new legal test is needed to address this problem, although new guidance will help to elaborate and apply the test to internet shutdowns.
We cannot point to a single internet shutdown that meets the test.
Provided by Law
We have identified at least 27 countries, out of 44 studied, with laws that could possibly allow shutdowns.3 This sample suggests that most countries worldwide have some law or regulation on the books that could be used to shut down networks. Yet the analysis does not end there. Laws must clearly alert the community to expected norms of behavior and the bounds of the government’s power to enforce those norms.4 General police powers and vague emergency and national security laws do not clearly instruct police on when control of telecommunications network is allowed, nor do they clearly delineate to ISPs and the public circumstances in which shutdowns can lawfully be ordered. The duration of shutdowns often appears arbitrary, and laws do not clearly state how long they should be expected to last.
Procedure matters, as well. When the laws are quickly formulated and passed, as in the case of Ethiopia's new powers, stakeholders including legislators do not have time to ensure adherence to constitutional and international law. Certainly, most shutdown laws were not passed by legislatures in the digital age, when our communications infrastructure has become essential to most daily functions. One regulation that was passed recently, the U.S. Standard Operating Procedure 303, remains cloaked in secrecy and unavailable to the public.5
Legitimate Aim
From our perspective, the aims of most shutdowns -- including to disrupt public protests, stop cheating on exams, or to slow rumors -- almost never fall within the three legitimate aims of Art. 19(3) of the International Covenant on Civil and Political Rights (ICCPR). These aims are to be “implemented narrowly,”6 and vague references to security or social cohesion do not satisfy the need for a clearly articulated, legitimate aim. Yet courts do not often enjoy the opportunity to apply this test. Officials ordering shutdowns either sit above the law and refuse to engage in judicial processes, or do not yet face challenges for disrupting communications networks. When it comes to shutdowns, procedural protections fail and impunity reigns.
Necessary and Proportionate
Shutdowns are neither necessary to, nor effective at, achieving a legitimate aim, as they often spread confusion and encourage more people to join public demonstrations. In the absence of more information about the actual harm that officials intend to prevent, the necessity of shutdowns cannot be proven.
The proportionality prong is not met: shutdowns are disproportionate as a rule, impacting everyone within the range of the blackout or on a targeted network or platform, not simply those who may be carrying out or contributing to a proscribed activity. That said, we expect governments, perhaps with the assistance of private companies, to improve their capacity to target shutdowns narrowly. Already, some governments have settled into patterns such as conducting “internet curfews,” where they shut down the internet only at night when the business community does not need it, or restrict only mobile and not fixed-line connections, which banks require to transact globally.
However, proportionality not only considers the number of individuals affected, but also the severity of the infringement of each individual’s human rights. The shutdown of a single cellular transmitter or tower in the vicinity of a public protest may impact only a small number of individuals; however the extent of the restriction on their freedom of expression -- an essential enabler of all other human rights in the digital age -- would be quite severe. This kind of insidious attack on expression would be more difficult to detect, appeal, and mitigate than the wide-scale restrictions that make global headlines, while still preventing the impacted individuals from exercising their human rights online.
Reference to the analysis of “blocking or filtering technologies” in the 2011 Report of the Special Rapporteur on the promotion and protection of the right to freedom of opinion and expression, Frank La Rue, is appropriate.7 Even where authorities do publicize lists of blocked sites, applications, services, networks, or platforms, digital and online platforms today offer two-way and multi-way broadcasting, so it is more than simply “access to content” that is being restricted. Rather, the inviolable ability to develop and form opinions, too, is directly infringed, along with protection of the right to life. Cutting access to these platforms can block access to emergency services, cut communications with loved ones, and restrict the free flow of information in times of conflict.
Why and when do government officials order internet shutdowns?
Having permissive laws on the books only partially contributes to a government’s decision to disrupt communications.
Justifications
There are outliers to every rule, and we do not claim to cover below every reason government officials use to justify shutting down communications networks. For example, judges in Brazil have prominently used their powers to block access to WhatsApp’s messaging services, in retaliation for perceived contempt of court by its parent company Facebook. However, these categories8 sum up the majority of cases we have recently documented:
National Security
National security may be the most frequently cited justification for internet shutdowns globally. It’s also the broadest and most vague. In Pakistan, the government has made localized mobile network shutdowns justified on these grounds a standard practice,9 cutting people off from the network during Islamic holidays such as the day of Ashura, while citing concern about terrorist attacks. Orders for the shutdowns are meant to cover a relatively small area, but it usually becomes necessary to “shut down a wider area than specified,10 to ensure there is no ‘spillover’ of service,” according to the Institute for Human Rights and Business. This means that the majority of mobile internet users in Pakistan are often affected -- a clearly disproportionate impact. State authorities have also cut off the internet for “security” during a wrestling event11 and Independence Day celebrations, drawing into question the legitimacy of the aim and necessity of the restrictions.12
Elections and political events
Elections have historically been a common precipitator of shutdowns. The government ordered a disruption in Iran during the 2009 elections.13 In 2010, Myanmar14 repeatedly shut down the internet in the months surrounding its first elections in 20 years.15 In October of 2015, officials used local government elections to justify a day-long shutdown in Garo Hills region in India.16 Worryingly, there has been a series of shutdowns during or around elections17 over the past two years in countries in East and West Africa,18 with governments disrupting the internet, cutting access to social media, or carrying out full blackouts in Chad,19 Togo,20 the DRC, Congo-Brazzaville, and Uganda.21 Yet the trend has spread to Europe as well, with Montenegro recently shutting down communications tools during an election.
Visits by government officials also afford governments with excuses to shut down communications networks. Authorities in the Indian state of Jammu and Kashmir blocked mobile internet during Prime Minister Modi’s visit to the Kashmir Valley.22 The people of Kashmir have experienced more shutdowns23 than those living in any other state in India, but this may have been the first time that justification was tied to a high-profile visitor. Similarly, in the Philippines, the only shutdown we recorded took place in 2015 during a visit by Pope Francis. It was supposed to be a targeted disruption of the area along the Pope’s travel route, but other areas were impacted to prevent spillover of signals.24 And in May 2016, the Vietnamese government blocked Facebook25 when U.S. President Obama visited.
Protests and demonstrations
Governments that want to quell dissent or stop protests use internet shutdowns as an extreme form of censorship, often while claiming that they are working to restore order or keep people safe. Many governments did not learn from Egypt’s experience in 2011, when its internet blackout drove more people into the streets.26 Countries that have used shutdowns to control information during protests include Bahrain,27 Cameroon,28Chad,29 Democratic Republic of Congo,30 Ethiopia,31 India,32 Iran,33 Kazakhstan,34 Libya,35 Mali,36 Myanmar,37 Sudan,38 Syria,39 Uganda,40 the United States,41 and Venezuela,42 among others.
School Exams
An increasingly popular form of shutdown aims at stopping students from cheating on exams. The first one that we’re aware of happened in 2014 in Uzbekistan during university entrance exams. At the time, the government did not admit that school exams were the reason for the shutdown, instead citing “urgent maintenance work on telecommunications networks.”43 After that, in June 2015, Iraq shut down the internet44 for national exams, and did so again one year later.45 In 2016, a full-blown trend emerged, with governments in India,46 Algeria,47 Ethiopia,48 and Iraq each turning off communications networks during national exams. Often in these cases, the entire nation loses access to the internet for the duration of the exam -- usually around three hours, sometimes repeating for a few successive days.
While cheating may well be an intractable problem in these nations, we are confident that the purpose of these restrictions does not rise to the level of a legitimate aim under Art. 19(3) of the ICCPR, and that there are more proportionate tactics available to education administrators to combat the proscribed activity. While harming rights and economies, these shutdowns also send a message: the nation’s access to the internet has such little value that it should be subverted for mundane purposes. We submit that this message conflicts with the education system’s goal of preparing youth to participate in the information society. Further, we also promise not to protest if proctors take away students’ phones for exams.
Legal gaps
Access Now has found governments are more likely to order internet shutdowns when laws are outdated or overbroad; when laws are not transparent; and when international standards do not clearly disallow shutdowns.
When there are outdated laws and overbroad definitions
Many of the laws that governments use to take over communications networks or impose shutdowns are ambiguous and outdated. For example, in India, the government can use a telegraph law dating from the end of nineteenth century (1885) to justify its decision to take over a provider’s network - or otherwise use an older, colonial era broad power on actions to uphold public order provided for in the Code of Criminal Procedure. In other countries, governments can use telecommunications laws that were passed or updated several years — or even decades — ago, when no one could have known how the internet would develop or impact economic growth and development.
In addition, there are some countries that have very broad definitions for a “national emergency,” which leaves the door open to abuse and misuse of the law. In the Democratic Republic of Congo, for instance, the Telecommunications Framework Law allows the government to ban the use of “telecommunication facilities, in full or part, for any period of time, as it deems fit, in the interests of public security or national defence, the public telecommunications service, or for any other reason.” As we have seen in the Democratic Republic of Congo,49 the Central African Republic,50 and Ethiopia, whose executive recently issued a State of Emergency declaration,51 the government can use these powers to legitimize cutting off internet connectivity.
The laws on the books often do not take into account new realities, such as the central importance in our lives of mobile communication technologies. Throughout the world, laws that allow blocking, shutting down, or taking control of internet access, including blocking specific apps or services, should be reformed, and their scopes narrowed in accordance with human rights law.
When local laws and processes lack transparency
In some countries, there is a troubling lack of transparency surrounding laws and processes that can lead to an internet shutdown. In the United States, for instance, the Electronic Privacy Information Center has been at the forefront of a legal battle to gain more information about the federal government’s mobile network shutdown policy, Standard Operating Procedure 303.52 The fight for transparency became more important after authorities ordered a shutdown at a San Francisco railway station during public protests.53
Similarly, in Italy, the prime minister issued a secret cybersecurity decree dated January 2013 that established a system of “ad hoc agreements” between the government and the service providers.54 Under the terms of these ad hoc agreements — a worrying prospect that erodes internet users’ privacy — a provider could be required to hand over control of its network to an intelligence agency in the interests of national security. This kind of secret bilateral agreement between a government and a provider does not uphold the rule of law, and fails another prong of the freedom-of-expression test: that restrictions be “provided by law.”
Opaque agreements like this are likely to lead to shutdowns and service disruptions without notice or accountability.
More often, however, there is simply no grounding in law. In Ghana, no reference to law is found in several threats issued by the police chief to shut down social media during the election. Rather, he cites the internet’s destabilization as reason enough for a shutdown.55 Likewise, in Uganda, the executive did not appear to reference law, but cited safety and public order as the basis for the shutdowns.56
When international standards and legal instruments allow them
The International Telecommunication Union (ITU), a major intergovernmental agency responsible for setting technical telecommunications standards as well as norm-building, has provisions in its constitution57 that could be interpreted to justify an internet shutdown. Article 34 on the Stoppage of Telecommunications gives license to ITU member countries to block telecommunications “which may appear dangerous to the security of the State or contrary to its laws, to public order or to decency,” while Article 35 on the Suspension of Services gives member states “the right to suspend the international telecommunication service.”
Whether or not these ITU provisions are ever invoked, their existence tacitly approves government decisions to shut down networks. The upcoming ITU Plenipotentiary Conference in 2018 would be a good opportunity for countries to amend the ITU constitution in line with human rights law and norms. Any provisions that allow blocking or shutdowns should be excised to accord strictly with human rights frameworks.
Telecommunications regulators also require internet and telephony service providers to sign restrictive licenses. Though these licenses are not readily available -- an area to push for more transparency -- telcos often cite them as a justification for their compliance with government orders to shut down networks and block applications and services.
Focus on India
Indian users have suffered broad use of shutdowns, with at least 22 in the first nine months of 2016.58 Most were carried out by state government agencies, often under the terms of broad legal powers — such as Section 144 of the Code of Criminal Procedure — derived from 19th century British Raj era provisions meant to empower police units and district administration officials to enforce curfew and confiscate property. When responding to protests or other public developments, many police officials and district administrators across India are making it part of their standard operating procedure to use this vague legal provision to issue orders to telecom providers to suspend mobile internet access across districts, and sometimes the entire state.
The Telegraph Act — via Section 5(2) — provides for a more specific legal power to restrict or otherwise interfere with the transmission of messages on the direction of the Union or state governments. But this provision lacks defined procedure when it comes to internet shutdowns. This very section was originally used to justify the tapping and interception of phone calls, until the Supreme Court in 1998 held it would be unconstitutional if the government did not specify additional, clearer rules on the process of issuance and safeguards regarding such interception orders. It would be reasonable to expect a similar treatment for the power to suspend internet services.
The Information Technology Act, India’s principal information technology sector law, has provided for another tool, even if not a perfect one. Instead of a larger suspension of the internet itself, it provides for a proportional, limited power in favor of the Central government — and state government officials in emergency cases — to issue individual web content blocking orders when certain grounds are met.
|
