SPECIAL COLLECTION OF THE CASE LAW ON FREEDOM OF EXPRESSION I. Introduction The state of freedom of expression online is substantially shaped by the moderation of content on dominant social media platforms. Their private community guidelines are the legal framework for discussions in the digital realm. Effectively, they constitute global law applying to billions of individuals, set by a few companies of the Global North. This raises fundamental questions on how to deal with private power. The answer varies substantially, depending on the legal culture: Some, in particular the United States, emphasize the principle of private autonomy. When users choose a particular platform, they accept its terms and conditions. If they are unbearable for the users, they can choose another platform. Thus, it is not on the state to impose any rules on this contract of private, autonomous parties. Others understand the issue of private power as an antitrust problem. Too few companies and quasi-monopolistic structures hinder the free market. The lack of competition, reinforced by network effects, leads to walled gardens, where consumers are left with no other choice than to pick one of the few existing networks and accept the rules of a “benevolent dictator”. In consequence, only breaking up global social media companies or, at least, setting up interoperability obligations for them could improve the situation. A third school of thought addresses the accumulation of private power as a human rights problem. While nation-states are the main and “classical” addressees of human rights obligations, powerful private companies are bound as well, although to a lesser degree. This approach is the one analyzed in this Special Collection paper. Claimants all around the world have initiated legal actions based on the assumption that a particular act of content moderation—such as the (non-)removal of content, account’s suspensions or deletions, among others—, violated their human rights. However, the application of human rights to content moderation does not necessarily facilitate its exercise in practice. “Chilling effects” as a result of too restrictive laws are well-known to the human rights discourse. Individuals alter the way they exercise their rights when they fear sanctions. It is not even necessary that a concrete action is in fact punishable, it is enough for individuals to be uncertain of the consequences they would face, to stop them from doing it. This could be understood as a call to moderate content as little as possible to enhance freedom of expression. But it’s not that easy. For example, hate speech often attacks a person’s dignity, which is also protected by human rights law. Beyond that, not only “chilling effects” but also “silencing effects” can be observed. Humans, in particular women and members of vulnerable groups, are leaving the online discourse because of the hate and aggression they experience there. Participation in the digital public space—mainly made up of social media platforms— lies at the core of what freedom of expression should enable people to do. Thus, the balancing of different rights is a difficult challenge for courts that review content moderation decisions taken by social media platforms in the first place. In the process of elaborating this paper, more than a hundred cases on content moderation from all around the world have been considered. Not all of them have made it into this paper. Due to the broad scope of the topic of content moderation, some limits had to be drawn: This paper does not include cases in which individuals fight over the legality of a particular statement. While such cases often take place on social media platforms, they do not concern content moderation in a narrow sense. They do not delve into platforms’ obligations to remove (or reinstate) content, rather –just like in an analog environment– they concern general issues of defamation law and freedom of expression. Also, copyright cases are not included. They do address the obligations of social media platforms to keep their platforms free from illegal content. However, their legal regime is highly specialized and outside the scope of inquiry of this paper. The “classical cases” of content moderation dealt with in this Special Collection concern the platforms’ handling of hate speech, political issues, defamation, and disinformation which are, due to a lack of a special legal regime, in most cases solved by applying general principles of law. Not only social media platforms (in a narrow sense) are exercising such content moderation. While 6