III. Discussion

After introducing readers to his extensive enquiry into platforms and regulation, the second part of the book links the scholarship on the regulation of Internet intermediaries with the ongoing discussion of constitutionalising Internet governance. Different themes subsume the idea of digital constitutionalism. At the base level, Suzor pursues three objectives: legitimising the power of major platforms; protecting the autonomy of technology; and promoting platform adherence to international human rights standards. To be sure, in line with the sociology of constitutions scholarship, Suzor depicts constitutionalism as a way of setting a social order in a world of new private and transnational powersFootnote ³ that has acquired social and constitutional relevance and has dispelled the public and private dimensions of law.^{Footnote 4} The prowess of the book is remarkable, given the difficulty of discussing constitutional solutions to such a complex scenario. Conversely, the risk of normalising the language of constitutionalism when addressing a context that is not characterised by its traditional (and normative) standards should not be neglected.

Suzor’s view of constitutionalisation can be understood as an application of Teubner’s civil constitutionFootnote ⁵ and, accordingly, can be subjected to the same scepticism.Footnote ⁶ Indeed, Teubner’s “Constitutional Fragments: Societal Constitutionalism and Globalization” provides a way to explore the possibility of producing law outside the domain of public law and inside private sectors of society, such as (in this case) the governance of the Internet. Teubner’s work is forward-looking: it considers how law reflexively develops within different private organisations. Naturally, Teubner’s theory has triggered many debates. An engagement with this strand of scholarship would have sharpened the arguments of the book, as well as advanced the discussion on constitutionalism and private forms of governance.

However, whilst Suzor reproduces Teubner’s model of societal constitutionalism, there are also some substantial differences. Teubner draws on Luhmann’s systems theory, which deals with complexity and differentiation in society and relies on a post-humanistic approach. In his analysis, Suzor does not follow the language of differentiation, and instead adopts a strong human-centric perspective. The result of these combinations is that law is not explored for its social function. In addition, Teubner devotes much attention to the constitutive side of constitutionalism. Constitutionalism has a double face: it is a limitation of power as much as its institutionalisation. The book does not problematise this latter aspect, and thus distorts the narrative towards an idealistic view of self-regulation and the idea of an “open, vibrant, flourishing internet”.Footnote ⁷ This is important because one of the major threats to having a diverse and plural Internet comes from the power of a few companies that the book attempts to legitimise. Teubner’s constitutional fragments theory promotes the autonomy of each regime, and in order to counterbalance the risk of having a private entity “subject its own expansion to self-limitation”, it foresees the role of “constitutional irritants”, which are massive external pressures exerted by non-governmental organisations, courts and politics. In short, the main matrix in societal constitutionalism is that there should be no direct regulation, but rather societal pressure on how constitutionalisation is developing.

Suzor starts his analysis with the idea that platforms so far have not been required to act according to the rule of law: they exercise power discretionarily and without accountability.

These regulatory loopholes (ie the lack of specification regarding decision-making) have led governments to exert pressure on how platforms “govern”. Suzor reminds us that “We are ceding control to them”, which means that we are witnessing an increasing delegation of public functions to Internet intermediaries. Consistently, Suzor argues that digital platforms need to exercise their power legitimately. There seems to be a mutually reinforcing narrative (the more a company is perceived as legitimate, the less troublesome it is to outsource public functions), which is more generally symptomatic of the intensification of privatisation and globalisation processes, whose pressure on traditional legal and political structures is difficult to contain.

Furthermore, Suzor maintains that, because governmental regulations “are coming” (eg to tackle fake news, hate speech and so forth), “There are opportunities here for companies to avoid regulation being imposed from outside by constitutionalizing their own operations”.Footnote ⁸

He stresses:

Suzor’s concern goes to “badly designed law” that will require intermediaries to police the content. Before further detailing the rest of the arguments, a first reflection should be made: self-regulation has not worked so far. While self-regulation better serves Internet intermediaries’ governance, the problem is whether it also serves societal interests. This is not to say that regulatory efforts from the private sector should be dismissed. Nonetheless, since the 1990s, the idea that governmental intervention would have undermined the development of technology has been dominant. This laissez-faire approach has become regulation by other means (ie that exercised by the platforms), with the consequent reshaping of the traditional legal framework. Julie Cohen highlights how network information technologies alter the future of law by transforming our fundamental legal guarantees, which is where legal rights and obligations are defined.Footnote ¹⁰

Moreover, private companies, for better or worse, have been active in developing their own standards and have generally anticipated regulations from states. The latest example is the banning of political advertisements by companies such as Twitter, Snapchat and others. Nonetheless, Facebook, the most important of all in terms of its capacity to sell advertisements, has until recently refused to commit to banning political adverts. As a matter of fact, Suzor reports cases in which digital companies themselves were not able to live up to their own promises.Footnote ¹¹

Digital constitutionalism is supposed to allow companies to set procedural limits on rule-making and enforcement by themselves to avoid arbitrariness. While agreeing that implementing procedures is an important step in opening the governance of private actors, it is unclear why self-regulation should be seen as a form of constitutionalism. The book assumes a rather formal approach in this regard. While Suzor dexterously touches on all of the main formulas of constitutionalism (eg rights, due process, rule of law), a critical engagement with those terms remains in the background. Although this argument reflects an optimistic view, following it uncritically may instead leave aside some important societal consequences arising from the proposed constitutional model.

For instance, Suzor does not confront the problem of concentration of power “as such”. He discusses how market forces influence platforms,Footnote ¹² but conversely disengages from the influence that these platforms exert on the current market structure and their operation under the logic of the market.Footnote ¹³ While Suzor attempts to solve the problems deriving from major digital platforms that have become central “nodes” within the Internet, he largely refrains from challenging their position. The solutions offered are devised “within the internet intermediaries”. How ought self-constitutionalisation to answer the problem of structural inequalities that come with the platforms’ control over most data flows? If one wishes to discuss constitutionalism substantively and look at rights, equality cannot be limited to content moderation practices.Footnote ¹⁴ The network capacity of these platforms profoundly affects different political economies and creates a situation of dependency for small actors on dominant platforms, consequently affecting online diversity and competition. Furthermore, the concentration of power in a few big companies triggers a domino effect on constitutionally protected rights. In other words, the more control a few platforms exercise over the Internet, the more does surveillance, profiling and microtargeting become ubiquitous and sophisticated, with potential repercussions for misinformation and propaganda. If we agree that equality is a constitutional principle, it would be prudent to treat the network capacity of these platforms as the root of the problem. By putting aside this aspect, we risk merely proceduralising a neoliberal situation.

Even if one wants to follow Suzor’s line of thought and read constitutionalisation from within the platforms, thorny questions are still present. One problem relates to the autonomy and adequacy of law. With state regulations threatening the autonomy of companies on one side of the coin, we must also pay heed to the other side, where big companies affect the autonomy of law and impair the larger project of self-constitutionalism. Briefly put, what is the role of law within each Internet intermediary? How will these self-adopted procedures enable users to challenge digital platforms’ operations meaningfully? Will these self-limitations be adequate to the function that they claim to have?Footnote ¹⁵ At its most basic level, the self-adoption of law means that law is not separated from the power that it should limit. Although it is a given that companies will develop, implement and enforce their own rules, it is equally predictable that self-adopted procedures will facilitate the digital company’s operation and policies.Footnote ¹⁶ The pertinent question then becomes how a basic model of constitutionalism can prevent the law from becoming an instrument for those in power.Footnote ¹⁷ While one would tend to think that major companies will start converging towards administrative law techniques to seek legitimisation, the challenge is to assess what the consequences are of major companies applying the language of the rule of law in their realms. Counterintuitively, it risks establishing a constitutional façade to companies and legitimising their legal forms.

To overcome some of the above-mentioned problems, Suzor argues that we need meaningful transparency on decision-making procedures.Footnote ¹⁸ He then moves on to international human rights, public shaming and organised social pressure as tools to harness the limits of self-limitation.

Another relevant aspect to consider is how the utilisation of human rights language may give rise to an elitist form of governance. The book claims that, in order to achieve long-lasting changes in how companies behave, there must be a consensus (presumably global) on what to expect from tech companies, because isolated controversies are ineffective and companies are subject to different requests from different stakeholders. Because human rights are universal, Suzor argues that it is possible to agree on some “clear and relatively unified demands” coordinated and presented by civil society. On this point, it is perhaps worth mentioning the contribution of global administrative law (GAL) studies, which highlights how difficult it is to agree on values (eg rights and legitimacy), and hence it is often easier to compromise on procedural aspects. While societal constitutionalism attempts to solve the strong functionalist nature of GAL through the language of reflexivity and constitutional irritants,Footnote ¹⁹ practical questions remain because of the “risk of capture”Footnote ²⁰ discussed above.

Furthermore, one might consider that the normative dimension of rights evolves because of the different tensions concerning their interpretation. Every conflict about rights enables a discussion of the scope of rights. It is not consensus that provides solutions for our rights on the Internet, but rather a constant debate on their contested nature.

Instead, Suzor seems to suggest establishing a standardised understanding of rights upon which companies could agree. The danger is that the language of rights might be twisted in favour of the companies’ business model. It should be borne in mind, as Suzor himself acknowledges, that different countries have different rights traditions (eg the difference between freedom of expression in Europe and the USA). As a result, a rights approach that is translated into big companies’ policies is likely to affect each country’s constitutional framework differently; although rights are understood as universal, their application is subject to different interpretations.

While the importance of enhancing the procedural dimension, and the need for transparency and accountability can certainly be endorsed, any debate about digital constitutionalism can nonetheless not avoid critical issues related to structural inequalities and the privatisation of law. The optimism of the constitutional moment evoked by Suzor deserves scrutiny. While conveying the language of constitutionalism, Suzor locks in an idea of autonomy and primacy of technology, which at the end of the day protects the neoliberal practices of big platforms and promotes managerialism in disguise.Footnote ²¹

Having said that, the breadth of the topics that the book impeccably covers is truly impressive. Overall, Suzor provides insightful, detailed and accessible reading for anyone interested in the regulation of Internet intermediaries, thereby paving the way for a much-needed in-depth debate on the future of the Internet. The book makes a novel contribution to the existing literature on Internet regulatory problems, especially because Suzor extensively digs into how platforms govern. Suzor provides the reader with an opportunity to meaningfully assess power relations within major platforms, and to become more conscious about our rights and how they are affected in the digital environment. Most importantly, Suzor shows that the time has come to move the conversation on platform regulation to the next level.