How Internet Shutdowns Amplify Dangerous Speech

On April 21, 2019, Easter Sunday, bombs went off in churches and hotels across Sri Lanka, killing at least 290 people. Within hours, the country’s government blocked access to Facebook, YouTube, WhatsApp, Instagram, Snapchat, and Viber, citing concerns that Sri Lankans would use the social media platforms to spread misinformation that could spark more violence. International human rights advocates criticized the move as violating freedom of expression. But internet shutdowns such as these threaten much more than the rights of the populace to express themselves. In impeding the flow of information, governments may in fact be amplifying Dangerous Speech rather than controlling it, leading to the possibility of more violence despite efforts to stop its spread.

What constitutes a shutdown and why do governments use them?

The term “internet shutdown” can refer to a wide range of activities. Usually, a shutdown is when actors – most frequently, the government – intentionally disrupt a population’s ability to access the internet. But within that general definition, there are many variations. Shutdowns can be widespread in scope, affecting an entire nation, or they can be specific and targeted, such as cases where individual towns may lose their ability to access or use the internet. Governments may completely shut off the internet so that the population cannot access any sites on any devices, or they may engage in partial shutdowns where only specific sites are blocked or slowed down in a manner that renders them practically unusable.

In 2018, Access Now, an NGO focused on maintaining an open and free internet, documented 188 internet shutdowns, up from 108 in 2017 and 75 in 2016. Governments frequently cite concerns related to public safety, the spread of rumors that could lead to violence, or even school exams as justifications for blocking access to social media or the entire internet. In some cases, such as the Sri Lankan example mentioned above, fears of rumors and violence do seem to be the actual motivating factor behind internet shutdowns. In many others, however, governments use such blocks to quell social protests or silence opposition voices before elections. Around the world, internet shutdowns have largely become a tool used by those in power to stay in power.

Government leaders who impose shutdowns out of fears of misinformation and violence point to the speed through which false information can spread on social media. Between January 2017 and June 2018, for example, 33 people were killed by vigilante mobs in India following rumors that circulated on WhatsApp suggesting that men were coming to villages in order to kidnap children, a topic that we have written about previously. And false information can indeed spread very quickly on social media, and this can have horrifying outcomes. But shutting down the internet may actually make violence more likely. This happens in two ways:

Shutting down the internet amplifies the power of Dangerous Speech by creating an information vacuum.

Following the 2019 social media block in Sri Lanka, it was difficult for Sri Lankans to get information about what was happening, who was affected, or who was responsible. The shutdown did nothing to stop people from trying to answer these questions, however; it just moved these conversations off social media. Authoritative sources were more difficult to access, but people were still hungry for information. In this context, rumors flourished.

Yudhanjaya Wijeratne, a Sri Lankan data scientist and social media researcher, was among a small group who volunteered as fact-checkers in the days following the incident. Writing for Slate, Yudhanjaya recalls being asked to verify rumors suggesting that the police were arresting those who used virtual private networks (VPNs) to circumvent the social media ban, that the water supply in Colombo had been poisoned, and that there were trucks driving up and down the streets of the capital city carrying unexploded bombs. False information also circulated about who was responsible for the attacks, with some suggesting that it was carried out by Sinhala Buddhists or that the attacks were part of a resurgence of the Liberation Tigers of Tamil Eelam, a Sri Lankan militant group that had been responsible for violent action in the country between 1976 – 2009. Still others blamed Muslims more generally.

In the chaos following an attack, rumors like these can quickly become Dangerous Speech. Messages suggesting that particular groups were responsible for the violence meet an audience that is still on edge following the gruesome and widespread attacks, and therefore might be looking to retaliate. Without access to their normal trusted sources of information, people may be willing to trust sources that they normally would not because it is the only “news” available. Also, people cannot quickly fact-check reports of continuing threats and therefore may feel that they are facing an imminent danger that would warrant the use of violence. Internet shutdowns therefore quickly become a factor amplifying the power of Dangerous Speech to persuade an audience.

A recent study by Jan Rydzak of Stamford University supports this possible causal relationship. Examining the impact of internet shutdowns on collective actions (protests, riots, etc) in India, the study found that although non-violent protests tend to lose momentum after an internet shutdown (possibly because they generally require more coordination), violent collective action actually increases.

Blocking public access to the internet impedes the ability of counterspeakers to respond to Dangerous Speech.

Around the world, counterspeakers play an important role in responding to hatred, fact-checking dangerous narratives, and making space for multi-sided conversations about difficult topics. When rumors abound in the midst of violence, it becomes especially important to ensure that pathways remain open for authoritative local voices to counterspeak against Dangerous Speech. The internet is uniquely conducive to counterspeech, as information can be fact-checked quickly, and it is possible for less powerful voices to challenge the prevailing narratives that may dominate other media sources such as television or radio.

Internet shutdowns therefore make counterspeech difficult. When there is a total shutdown –as when Houthi rebels in Yemen damaged fiber optic wires, cutting off internet access to 80% of the country, there is little that can be done to assist online counterspeakers. If the block is partial, however, and access has only been blocked to particular sites, online counterspeech may still be possible. Using a VPN to circumvent the shutdown and access social media is one of the most popular methods used by counterspeakers to resist partial internet blocks. On July 14, 2018, for example, the Iraqi government disrupted access to social media amidst large-scale anti-government protests. Less than a week later, Psiphon, a very popular free VPN service, had grown from having just a few users in the country to having over four million. But even if those wishing to counterspeak can go through the effort to download and use a VPN, not everyone will be able to, or know how, to do this, limiting the potential audience. Therefore creating resources to educate populations about how to use VPNs, and making trustworthy VPNs easily available and free of charge, should be a major goal for those wishing to counter the Dangerous Speech that occurs during internet shutdowns.


Those opposed to internet shutdowns regularly reference harms such as violations to freedom of expression, inability to contact family members during a crisis, and the ways that social media access can assist in humanitarian rescue efforts. Recently, advocates are also focusing on the economic impact of shutdowns, an argument that has been particularly appealing to government officials.

But emerging studies on the causal relationship between shutdowns and violence and an understanding of how Dangerous Speech functions suggest that the possible harms extend even further. When people cannot verify or refute rumors, and cannot access alternative narratives or trustworthy counterspeech, there is an increased likelihood that these rumors will become Dangerous Speech. Therefore, if governments are serious about addressing rumors that could inspire violence, they must resist the urge to shut down the internet.