As mentioned above, the UN Declaration of Human Right states that “everyone has the right to freedom of opinion and expression; this right includes freedom to hold opinions without interference and to seek, receive and impart information and ideas through any media and regardless of frontiers.” This is freedom of expression in a broad sense, as the wording contains both individuals’ and organizations’ freedom to express themselves and their right to seek out information freely and search for information freely.

In today’s online world, freedom of speech is under pressure, because the tech giants have been given a position where they can lay down the rules of public conversation and access to information. This is a consequence of freedom of expression having been capitalized by tech giants. The right to express one’s opinion and seek out information is no longer exercised without interference, because the tech companies adapt and limit the free information exchange of individuals and organizations, based on commercial interests as well as supercilious, paternalistic ideas. First of all, and as discussed in previous chapters, interference takes place, and it takes place automatically and opaquely through the platforms’ algorithm systems. But there is also a manual, censoring interference taking place, based on policies formulated by the giants in order to create an information environment which aims to reflect their own ideals. These policies are vaguely and hazily worded. The phrasings reduce complex legal decision making, which we will elaborate on in the following chapters. This interference means that who and what is gaining ears today is increasingly determined by large private technological players.

There is no doubt that the creation and spread of the Internet initially caused a landslide towards increased real freedom of expression. Any user could now create their own website or blog from which to preach their incontestable opinions about nothing and everything and participate in many different debate forums and commentary threads and make serious or less serious comments about other people’s statements, positions and personalities. If you were smart or lucky, you could get in touch and debate with a brand new and broad readership—and moreover, you were exempt from the objections of opinions editors. In the classic media setup, the opinion editor was responsible for combing through written pieces submitted “to the editor” by ordinary people in order to decide which ones to promote and expose. Of course, one was still subject to the laws of one’s country and its partial restrictions on freedom of expression—although it has taken most states a long time to adjust, not only to monitoring the printed word but also the digitally communicated word.

The advent of the Internet also meant highly increased freedom of speech in the broader sense, the one about seeking out information: the ever-increasing data volumes available on the web allow the user to access news, search information, gain knowledge, get informed about debates, positions and controversies, all at an incredibly fast pace. The top tech representatives never fail to mention, in manifestos and speeches, how much they believe their companies contribute to this increased freedom of expression. In November 2010, at the doorsteps of the “Arab Spring”, Eric Schmidt and Jared Cohen from Google rejoicingly described how dedicated they were to “promote freedom of expression on the Internet and protect privacy.”Footnote 1 They even envisioned rebels equipped with cell phones able to summon “flash mobs” and “... shake repressive governments, building new tools to skirt firewalls and censors, reporting and tweeting the new online journalism, and writing a bill of human rights for the Internet age.”Footnote 2 It sounds like a Diderot or a Tom Paine of the twenty-first century—eloquent and ecstatic declarations of Enlightenment, hard not to be moved by.

But are these giants really defenders of freedom? As early as 2010, significant slippage was already well underway. The anarchist utopia of the early Internet, with its many individual players and their self-organized structure, had proven just as illusory as the free liberal market of classic liberalism with its multiple parties with equal status. Both phenomena share one and the same reason; in an open market, the best providers can invest profits to become even better and, in many cases, they can eradicate enough of their competitors to approach monopoly. Tech giants such as Amazon (founded in 1994), Google (founded in 1998), Facebook (founded in 2004) and Twitter (founded in 2006) became some of the world’s largest companies during the 2000s and 2010s. Not only did they thrive on the Matthew Effect—“for whoever has, to him more shall be given”—but also on the even more important network effect; the sheer size of the network is in itself part of its attractiveness, so that the largest networks naturally attract new clients who are eager to get in touch with as many other users as possible. At the same time, the marginal cost of adding new users is incredibly small. A new user, however, continues to be just as valuable to the company as when they boarded, because advertisers are willing to pay in proportion to the number of users reached, the number of times viewed and the number of clicks.

An important aspect to this attention economy is, of course, maintaining the users once they sign up. This is done by offering them useful, convenient and free services—but also by not scaring them away. Among other things, this means not showing them content they might find offensive. This basic trait means that from early on, the big tech companies formulated policies for content removal as part of their “user policies”, “terms of service” or “community standards”, on how the user was expected to behave on the platforms. If users did not comply with these rules, their posts could be deleted and their access blocked—temporarily or permanently. More often than not, these rules were only described scantily and in generic or hazy terms—but in all cases they were significantly tougher and more narrow than the legal restrictions on freedom of expression in most democratic countries.

This uncertainty is multiplied by the fact that individual tech companies have different versions of these rules. Oftentimes they will begin with a general version, the “terms of service”, which is the legally binding description of the mutual obligations of both service provider and user.Footnote 3 Secondly, a more detailed version is outlined in the community standards. And thirdly, a considerably more detailed version in the form of the internal staff guidelines, a document not available to the public and which is usually only known if leaked by a whistle blower.Footnote 4 For a long time, such control of what can be expressed on the tech giants’ platforms could seem completely unproblematic. They are private companies. In a sense, they run a new kind of media in mutual competition with each other. Their community standards then correspond to the editorial policies of newspapers. Thus, they correspond to the fact that—as a mere matter of course—we expect a difference between what can be published in TheNew York Times and in The New York Post. In a competitive market, private companies have the full right to set their course as they wish and exclude points of view they do not like to bring. No one’s freedom of speech is violated, if a newspaper turns down, removes or ignores a letter or op-ed, because people can always take their statements elsewhere—on the Internet, people even have abundant ways of making their own websites, deciding on the editorial line themselves.

The problem with such an analysis of tech giants as mutually competing media is that their enormous success of recent years has in fact rid them of real competition—they are increasingly bordering on de facto monopoly. In practice, Microsoft and Apple have a duopoly in the field of software, while Google and Facebook are on their way to becoming a duopoly when it comes to their core business: selling ads based on free digital services. As mentioned above, together the two have more than 80% of the American online advertising market and are only threatened by Amazon. But they are also monopolies seen from a user point of view: As of April 2018, Google had 90,61% of the world’s search activity—distant followers are Bing (3,24%), Yahoo! (2,09%) and the search engine of the Chinese government, Baidu (2,04%); all other search engines have less than 1%.Footnote 5 Google is estimated to have more than 2 trillion searches per year, which corresponds to 2.5 million searches per second. By 2016, Google’s parent company Alphabet had six other platforms, each with more than a billion users: Gmail, Android, Google Chrome, Google Maps, YouTube and Google Play Store. For some years now, Facebook has had between 65% and 85% of the social media market share; as of January 2017, it was 87,3%, dropping to less than 70% in the spring of 2018, possibly due to the Cambridge Analytica scandal of March 2018. Competitor numbers are substantially lower: Twitter and YouTube (owned by Google) have around 7% and 8% respectively, while Pinterest, who specializes in photo sharing, now has 15%.Footnote 6

Companies which have a de facto monopoly can easily overcome pressure from the competition. Getting new users is cheap for them; they can name prices themselves (in this case for advertisers), because no other players can offer such extensive market penetration; and they can continuously buy out hundreds of emerging competitors. In that way Google, among many other deals, have purchased Android (2005), Measure Map (2006) YouTube (2006), DoubleClick (2007), Ebook (2011) and Motorola (2011); their most recent larger acquisition is GIF search engine Tenor (2018). As part of a long list of other acquisitions, Facebook took over FriendFeed (2009), Friendster’s patents (2010), Instagram (2012) and WhatsApp (2014); the most recent large acquisition is Confirm (2018), an identification verifier.Footnote 7 Facebook has even developed a tool by the name of Onavo Protect, which is marketed as a service to protect user data but which in fact maps the online behavior of users; that makes it possible to monitor whether new apps or services are becoming popular among Facebook users. If that is the case, Facebook will simply acquire the new company, which is what, most famously, happened with WhatsApp.Footnote 8 In August 2018, however, the Onavo app was excluded from the App Store, because it did not comply with Apple’s policy for data collection and privacy. On previous occasions, Apple CEO Tim Cook has criticized Facebook’s approach to privacy, but this was the first time Apple actually acted upon it by excluding a Facebook app.Footnote 9 Later we shall discuss the possibility of subjecting tech giants to anti-monopoly regulation.

When companies such as Google and Facebook are approaching de facto monopoly status, it is no longer such an innocent matter what statements are prohibited, marginalized and removed from their sites, as a breach of their community standards. Especially for younger generations, services such as these have become the main source, for many the only source, of news and public debate. But regardless of how much their top managers say so, eloquently phrased when glasses are raised, the community standards no longer reflect an ambition of free debate, in the sense of inclusion of all points of view and protection of the freedom to also tolerate provocative, strange, and unpopular expressions. Rather, the community standards reflect an ambition of not scaring off users. And an ambition not to offend advertisers by showing their ads anywhere near controversial content. This puts them right between two opposing considerations: on the one hand the freedom of expression, which is often praised in odes to openness, transparency and access for all persons and points of views. And, on the other hand, a detailed policy for the removal of content, a policy which, however, does not enjoy the same openness but which generally is kept hidden, both when it comes to detailed guidelines and what motivates the practical interpretation of them.