Legal but harmful content and little foresight in supervision

Last week I had the honor and pleasure of participating in the conference organized by Columbia Global Freedom of Expression Program: Regulating the online public sphere: From decentralized Networks to Public Regulation. The conference, which had more than 500 people connected online listening to the panels, seems more relevant than ever. During the conference, UNESCO announced a meeting to be held in February 2023 to discuss a regulatory framework for social media and cites, among other arguments, the existence of more than 190 bills around the world that attempt to regulate platforms in one way or another.  

As Catalina Botero pointed out in her introduction as host of this event, we are seeing a battle around how to deal with expressions on the Internet. And there are different theories in this regard, which our host highlighted succinctly and clearly: 1) those who maintain that platforms can and should self-regulate in compliance with international human rights standards when implementing said self-regulation; 2) those who subscribe to the hypothesis that the platforms should be neutral towards the content of their users and regulated as telephone companies, without any liability for third party content, and even with the obligation to host the platform without any discrimination regarding the content (common carriers); 3) those who argue that the self-regulation of platforms is part of the right to free industry and that contract law should govern the relationship between users and platforms, between content and moderation; 4) those who equate platforms to the media and therefore advocate applying liability criteria as editors. I dare to say that several proposals touch on elements of all four positions, even when they are in direct and open contradiction.

A first observation that arises from the positions raised, carried over from the previous day when decentralization bills were discussed in the face of the problems of moderation, is that I do not believe that there is a consensus about what are the "problems" that we want to solve with these endeavors. Whether we are thinking of decentralization tools or the regulation of social media, it is difficult to compare the 190 bills, since I know that some (and, I suspect, many) pursue radically different objectives and even contradictory perceptions of the “problem.” For example: if we define “the problem” as the abuse of private censorship by the platforms, solutions like the one suggested in number 2 will make sense. If we define “the problem” as the proliferation of problematic speeches (a term that is beginning to appear in certain documents as art vocabulary, but whose definition is ambiguous and unclear in and of itself) as a response to a discretionary hand with disproportionate interference in the moderation of public debate, we will see solutions more aligned with number 1. If instead, we see "the problem" as an editorial problem -understanding hosting and moderation as possible acceptance of the content- the proposal in number 4 will acquire preponderance. If we understand "the problem" as non-compliance (or the need to enforce) with what was agreed, the perspective offered by number 3 would be perhaps the most accurate. In short, when we talk about proposals to solve content moderation problems, we should bear in mind that the definition of the problem is not unequivocal. Or that not all proposals define the problem in the same way - explicitly or implicitly. All of these are probably, to a greater or lesser extent ,possible current problems and we presumably cannot solve them with a single regulatory proposal.

Having read some of the 190 bills, including a good part of those that are regularly cited in our region (Latin America in general), I argued, in the panel that I had to be part of, that there are two important contradictions in the regulatory projects that we have seen so far. The first is that in the background of this discussion, in my opinion, lies a peculiar problem: what the law should do with any legal but "harmful" content or expression (Americans, as always, are creative and graphic and designate this distinction legal but awful, or harmful but legal). The second contradiction lies in the supervisory mechanisms or lack thereof. Over the next few weeks, I will upload some comments on the subject. I will start with lawful but awful. I will continue with the (possible) difference in the Latin American approach. The next one will be supervision (which I hope already comes with a link to a recent paper that CELE wrote). 

The "legal but harmful expression" category does not imply a contradiction in and of itself. Although it is a general principle of civil law that all damage must be repaired, there are exceptions. There are conducts, actions, and omissions that the law assumes as legal but harmful. For example, according to the state, smoking is legal but harmful. Drinking alcohol is legal but harmful. The tolerance that the law has against damage differs in each topic. Regarding an issue close to our reality, in the global jurisprudence of freedom of expression, we have universally accepted that public officials must have more tolerance against criticism and interference with their privacy. The Inter-American Court, for example, following the European Court, explains that the distinction is not for the person but for the social function they fulfill in a democratic society. Cases of defamation and slander against public officials rarely prosper because transparency and democracy require a vigorous public debate about the exercise of their work, ethics, and integrity. There are potential damages to the honor and privacy of public officials that the law mandates to tolerate, and as such, are not susceptible to reparations or subject to any sanction, except in exceptional cases (actual malice). To the same extent, the Inter-American Legal Framework already establishes that the right to freedom of expression not only protects conciliatory and innocuous expression but also expressions that "offend, are shocking, or disturb the majority." 

When content moderation problems are popularly discussed, the debate has focused on legal but harmful content. Some of the most abundant examples in the latest bills are disinformation, misogynistic speech, problematic speech, radicalized speech, polarizing speech, offensive speech, and violent speech. All of these, as defined in recent bills, are legally protected. There is no legitimate reason for the state to limit them. Moreover, there is a legitimate reason for the state not to limit them. 

There may be several explanations as to why the focus is on legal but harmful speech. I suspect, and this is what I am writing with Paulina Gutierrez about, that the adoption of section 230 of the CDA can somewhat help understand this priority: this offers immunity for companies for third-party content and also for the moderation of said content. Something similar, although not identical, happened in Europe with the Electronic Commerce Directive. Our theory is that section 230 removed the conduct of the platforms from the legal realm. Moreover, the jurisprudence after the adoption of section 230 extended this immunity to the point that nothing that the companies do with that third-party content can even be analyzed by the law. The immunity is such that until now a large part of the demands has been outright rejected, without even a possible analysis. This, which at first made the internet flourish, today makes the actions within the different services that the platforms provide (ordering, indexing, showing, hosting, moderating, recommending, making visible, hiding, and many other actions ) are, in the collective imagination, outside the sphere of the law and left exclusively to the sphere of ethics, morality, and traditions, or corporate social liability. 

In the beginning, when companies in the sector executed a few actions (hosting, ordering, indexing, and a little more), this system made sense. But now, with the power Internet companies, particularly social media, exert to inform, shape, limit, or fuel public debate, immunity seems complicated. Modern regulatory proposals emerge in this scenario: some focus on content, but also others focus on processes. 

And here I have a second observation: the distinction between legal and illegal content is made by the state. Until now, the distinction between ethical and responsible content was made by the platforms. However, the vast majority of the current regulatory proposals for social media are not neutral in the face of lawful content. What underlies is a state concern for lawful speech and a desire to impose new limitations on freedom of expression from a perspective of ethics or corporate social liability. A large part of the bills and laws that we see today, including those that come to regulate processes, imply some renegotiation of the legitimate limits to freedom of expression on platforms. And insofar as what is regulated is how the expression circulates, this renegotiation is done indirectly: putting on the company the obligation to self-regulate to avoid damages in a broad and not strictly legal sense; establishing obligations for companies regarding the lawful but “problematic” speech of their users. These damages, according to our substantive laws, are lawful, or at least should be tolerable. 

The "problem" with content moderation in general is not illegal speech. Illicit expressions are regulated, whether the company that hosts the content says so or not, since the limits to freedom of expression are negotiated in each state and each region according to their needs, cultures, history, and always according to minimum human rights standards. Individual responsibility for such expressions is subject to prosecution. 

The problem is the legal discourse. If the state cannot legitimately restrict it (except in matters of time, form, and place), it cannot order a third party to limit it either directly or indirectly. In these cases, perhaps, it is worth seriously debating whether the current permissible limits to freedom of expression are adequate. It is also worth addressing the legal analysis of new phenomena, such as virality or permanence. What was previously tolerable according to the law because it was not permanent or did not have greater visibility, today might not be so. In any case, this is a more honest debate, fundamentally, about permissible limitations to the right to freedom of expression. 

Now, there are legislative proposals that promote, in addition to these indirect limitations, other measures that could be welcome. Among them, multiple initiatives propose greater transparency about the actions of the platforms themselves. There is a consensus around the need for greater transparency. But there is no consensus on the need for transparency required or supervised by the state. Nor is there a consensus about transparency regarding its object and purpose. And here again, there are underlying problems with the distinction between substance and form. Depending on the answer to these questions, we will emphasize some pieces of information and not others. The objective pursued by transparency will be decisive in making sense of it. Transparency must be understood as a means to an end. And the purpose must be legitimate, especially if it is the state that commands such transparency. According to what we pursue, we will ask for one piece of information instead of another. We will highlight some aspects and not others. We will audit certain sectors and not others. I fear that there is no great consensus here either, at least not in the regulatory proposals that exist today. 

Perhaps the most advanced bill of law in this regard is the DSA, and it also has problems regarding broad definitions and a lack of clarity about how said transparency will be effective. On the one hand, there are general obligations of transparency regarding the terms and conditions of service, moderation mechanisms and criteria, and aggregate information regarding the amount of affected content, affected users, decisions made, and follow-up of reports. On the other hand, it is established that the information supporting said reports may be audited, although it is not yet clear how. Finally, two additional transparency obligations are set up: one is the support of moderation decisions and their communication to users. The other, in the case of large platforms (more than 45 million users in the European Union) is the obligation to carry out a risk analysis of the products against potential damage to democracy, the integrity of minors, and the safety or integrity of users, among others. I return here to the central point of this piece: the risks identified by the proposal, which require transparency and action (risk analysis and the obligation to mitigate said risks) are based on legal, although potentially harmful, expressions. The details regarding what these risks mean, how to carry out these audits, the criteria to evaluate results, the indicators of success or failure of said initiatives, remained to be defined in the implementation phase, with the supervision mechanism that I will refer to in the next installment. <img draggable="false" role="img" class="emoji" alt="" src="https://s.w.org/images/core/emoji/14.0.0/svg/1f609.svg">