Germany’s Online Crackdowns Inspire the World’s Dictators

The German government has responded to a recent spate of right-wing extremist violence with a commitment to enact tough new measures against online hate speech. These would expand Germany’s existing efforts along these lines—but seemingly without any consideration of the pernicious effects they’ve already had around the world. As our new report shows, Germany’s ongoing crackdown on online speech has been closely watched, and copied, by authoritarian governments eager to curb political dissent.

The crackdown began after 2015, when Chancellor Angela Merkel’s decision to welcome more than 1 million refugees and asylum-seekers was greeted with a tidal wave of protest ranging from derision to visceral hatred on social media. The anger migrated from the online trolling and memes of keyboard warriors into actual attacks on refugee centers across Germany.

Sensing a loss of control, then-Minister of Justice Heiko Maas wrote a stern letter to Facebook warning the tech giant that, despite the need for free speech, “The internet is not a lawless space where racist abuse and illegal posts can be allowed to flourish.” Maas demanded much stricter policing of content violating German law and Facebook’s own community standards. Facebook and other tech companies agreed to a voluntary deal under which they would remove content deemed illegal within 24 hours.

Facebook’s removal rates failed to satisfy Mass, however, and he decided that legally binding measures were necessary to curb the flood of online hate. In 2017, the government introduced the Network Enforcement Act, which was adopted that same year and entered into force early 2018.

The law imposes so-called intermediary liability for social media networks with over 2 million registered users. Any content, which is “manifestly unlawful” must be removed in a time frame of 24 hours. For all other “unlawful” content, the deadline is seven days. Failure to remove illegal content is punishable by fines of up to 50 million euros, about $55 million.

The Network Enforcement Act’s detractors argue that it delegates to the private sector the role of cybercop with little transparency or due process. This encourages social media companies to over-implement by providing an incentive to err on the side of caution to avoid fines. While no new offenses were crafted for the law, some of the existing definitions of “unlawful” are problematic in themselves.

This includes most glaringly Germany’s approach to the offense of “defamation of religions,” which violates international human rights standards. Article 19 of the International Covenant on Civil and Political Rights guarantees freedom of expression including the “freedom to seek, receive and impart information and ideas of all kinds, regardless of frontiers.” In 2018, the United Nations’ independent expert on freedom of expression expressed concern about the new act and other state-imposed models of intermediary liability as potential violations of Article 19.

The Network Enforcement Act has woven itself into the landscape of internet governance at a time when many states worldwide are pushing for increasing regulation to fight online threats, both real and imagined. The critics of the law who warned that the act might legitimize a model of online censorship that can readily be adapted to serve the ends of authoritarian states have been proved right. Our new research shows that in less than two years the law has essentially been copy-pasted by governments around the world—most of which do not match Germany’s commitment to democracy, the rule of law, and human rights.

Since the adoption of the new German law, at least 13 countries—in addition to the European Commission—have adopted or proposed models of intermediary liability broadly similar to the act’s matrix. According to Freedom House’s 2019 assessment of freedom on the internet, four of those countries are ranked as being “not free” (Venezuela, Vietnam, Russia and Belarus, Honduras is not surveyed but is ranked as “not free” on press freedom), five are ranked “partly free” (Kenya, India, Singapore, Malaysia, and the Philippines), and only three are ranked “free” (France, the United Kingdom, and Australia). With the exception of India, Kenya, Vietnam, and Australia, all these countries—as well as the European Commission—have explicitly referred to the Network Enforcement Act as an inspiration or justification for their models of intermediary liability.

Moreover, several of these countries, including Venezuela, Vietnam, India, Russia, Malaysia, and Kenya, require intermediaries to remove vague categories of content that include “fake news,” “defamation of religions,” and “anti-government propaganda,” and many of them include overly broad definitions of hate speech that go much further than the German law. A Russian bill signed into law by President Vladimir Putin in March is a good example. It defines “unreliable information” as follows:

“Socially significant information disseminated under the guise of reliable messages, which creates a threat to life and (/or) the health of citizens or property, the threat of mass disturbance of public order and (/or) public safety, or the threat of creating or impairing the proper operation of vital elements of transport or social infrastructure, credit institutions, energy facilities, industry or communications.”

Any controversial opinion or criticism of the government could plausibly be covered by this overly broad and vague definition, which falls considerably short of international human rights standards.

The explanatory report of the Russian bill explicitly referred to the Network Enforcement Act, and, responding to criticism, Kremlin representatives argued that false information “is regulated fairly harshly in many countries of the world including Europe. It is therefore of course necessary to do it in our country too.” This is a clear example of how Germany’s internet law provides cover for authoritarian states attempting to restrict online content.

In May, Singapore adopted the wide-ranging Protection from Online Falsehoods and Manipulation Bill. The law includes a vague definition of “false statements of fact,” authorizing a minister to issue directions to internet intermediaries that must correct or disable content. Prior to the adoption of the bill, a preliminary report referenced the German law. What might constitute false statement of facts? A 2018 report by the policy forum of Singapore’s ruling People’s Action Party highlighted a Human Rights Watch report critical of press freedom restrictions in Singapore as based on “deliberate falsehoods” used to “advocate political change.” In other words, the law may well be used to target human rights and civil society groups shining a critical light on the Singaporean government.

However, both the Russian and Singaporean laws pale in comparison to the sweeping category of illegal content covered by Vietnam’s draconian Law on Cybersecurity, which passed in 2018 and prohibits: “propaganda against the Socialist Republic of Vietnam,” “distortion or defamation of the people’s administrative authorities,” “psychological warfare … causing division or hatred between [Vietnamese] ethnic groups, religions and people of all countries,” “insulting the [Vietnamese] people, the national flag, national emblem, national anthem, great men, leaders, famous people or national heroes,” and “invented or untruthful contents causing confusion amongst the Citizens.”

It is also problematic that some states, including Russia and Vietnam, have established (or proposed to establish) governmental entities responsible for notifying and ordering intermediaries to remove illegal content without any independent review or complaint mechanisms. When you combine the sweeping nature of the prohibited content with the lack of meaningful oversight, it is clear that these laws will serve to further entrench tight government control over an already restrictive online sphere.

Several of the laws or proposals we identified, including in the U.K. and India, don’t merely rely on a notification and takedown regime. They establish a “duty of care” requiring intermediaries to actively police and preventively remove illegal or undesirable content. Such a step may encourage automated moderation or filtering of user content through artificial intelligence and upload filters. This essentially poses the risk of reintroducing government-mandated prepublication censorship enforced by private tech companies. In the 18th and 19th centuries, European censors lost the struggle to keep up with clandestine publications and increasingly sophisticated printing presses. Consequently, pre-publication censorship all but died in the second half of the 19^th century, only to be revamped and revitalized by totalitarianism in the 20^th century. But in the digital age, censorship may ultimately be able to search and destroy undesirable content at a scale previously unimaginable.

All these developments suggest that the Network Enforcement Act has provided an important impetus for, and legitimacy to, models of intermediary liability that violate freedom of expression as set out in Article 19 of the International Covenant on Civil and Political Rights. This development contributes significantly to the weakening of the already perilous state of internet freedom around the world.

The adverse consequences of the Network Enforcement Act were certainly not intended by the German government. In conjunction with the German constitution and commitment to the rule of law, the act provides safeguards absent from the most draconian laws subsequently adopted by other states. Indeed, several of the countries that have mimicked the German law had already implemented severe restrictions of online freedom prior to the German initiative and would likely have tightened their grip irrespective of the Network Enforcement Act.

Yet, the act seems to have provided several states with both the justification and the basic model for swift and decisive action. Unwittingly, Europe’s most influential country has contributed to the erosion of global internet freedom by developing and legitimizing a prototype of online censorship that can readily be adapted to serve the ends of authoritarian states.

In a world where both online and offline speech is under systematic global attack, democracies have a special obligation to err on the side of free speech. Succumbing to the ever-present temptation of fighting illiberal ideas with illiberal laws is shortsighted and irresponsible. Once democracies cede the high ground, authoritarians will rush in, creating a regulatory race to the bottom. This entails severe and negative consequences for free speech, independent media, vibrant civil society, and political pluralism, without which authoritarianism cannot be defeated, nor democracy defended.