Technology creates societal fractures. Courts have traditionally been responsible to apply the law to fill fractures, ease normative tensions, develop social mores. But courts are uneasy rulers in times of technological change, even as platforms seek to recreate private quasi-courts. Yet their power – especially when mediated by automated tools – is substantial. In more ways than one courts are the hidden rules of digitality. High time to rediscover their powers.

by Matthias C. Kettemann, Wolfgang Schulz  and Giovanni De Gregorio
 
Be it international treaty-making, top-down or multistakeholder-based development of new norms: amongst state actors, governments are the most active players on the international Internet governance scene. National and European legislators are also prominently engaged in normative activities with regard to the Internet. Just consider, locally, Germany’s Network Enforcement Law (NetzDG) and, regionally, the proposals for a Digital Services and a Digital Markets Act.
 
But which institutions are missing? What institution has substantially impacted the EU’s data transfer practices? Which bodies developed balancing exercises between privacy rights and freedom of information for content that someone wishes to disappear from the Internet? Who ruled that certain intermediaries had to delete obviously illegal content without any delay – and even beyond borders, if international allows? Courts, courts and courts.
 
Largely unnoticed, a growing body of cases has emerged that – read together and analyzed separately – greatly influence the development of norms on the Internet and have substantial impacts on the rights of citizens-consumers. From Delfi v Estonia to Facebook Ireland, from Google Spain to the III.Weg, the role of courts in structuring the public’s use of the Internet and indeed the ‘public’ness of the Internet has been substantial.
 

Private “Quasi-Courts”

But courts are not alone in normatively structuring the Internet. Online platforms, especially social network services providers, have become their own quasi-courts. A growing number of moderator-judges takes thousands of content-related decisions each day. Helped by artificial intelligence they have become key decision-makers that shape what users are allowed to publish and read.
 
Faced with substantial critique as to the opacity of this process, at least one major social network provider, Facebook, has prominently installed a dispute settlement mechanism, an Oversight Board for content decisions. This auto-judicial approach has both its merits and its limits. While relying on internal quasi-judicial proceedings, Internet companies have not yet been too enthusiastic in joining other more state-oriented models of regulated self-regulation, as foreseen e.g. in the Network Enforcement Act. But voluntary self-control mechanisms can work, as the example of the procedures instituted by Germany’s multimedia services providers in the context of protection of young people show. 
  
Alternative judicial procedures within Internet companies can be more reactive and reach decisions more quickly. But many questions remain: What about the members, their selection, independence, qualification and time in office? How to ensure necessary consistency with other opinions, cultural sensitivity, commitment to values, transparency? These issues have been discussed with regard to traditional courts for decades; quasi-judicial institutions need to find convincing answers. Especially, as new forms of private Supreme Courts emerge, such as the Facebook Oversight Board whose role – an interpreter of rules? A rule-maker? – remain unsettled.
 

Hybridisation

This mix of public and private adjudication leads to a hybridisation of the very legal bases on which courts take their decisions. While the power of governments to create norms could be efficient but usually limited for separating powers, traditional legislative procedures are much slower. How do the norms on which judges base their decisions develop? How transnational organisations and businesses contribute to this framework? How can highly trained judges intervene meaningfully in the content moderation activities of social network services providers? And are judges always better than algorithms in discovering, for instance, patterns of discrimination that are premised upon big data analytics?
 
Before the rise of private adjudication or alternative dispute resolution systems, traditional courts, however, need to be asked similarly challenging questions: how do they situate themselves and their cases vis-à-vis other courts deciding on similar fact patterns? And what about the geographic reach of judgments? Are judgments enforceable across Europe? How do they engage with the possible consequences of their judgments when these might extend to areas outside of their jurisdiction? Is a new rule of comity between judicial systems emerging?
 
In light of the growing number of cases a new comparative approach seems to make sense. Closer cooperation between national courts and quasi-judicial institutions could also be a solution, just as the increasing use of algorithms in courts and the judicial control of algorithms in use by companies and alternative dispute mechanisms. But we can only draw the necessary conclusions based on case data. The Global Freedom of Expression case collection at Columbia University and the database of the Internet & Jurisdiction Network are important resources, but a more thorough analysis is necessary.
 
This analysis might also find out which judicial and quasi-judicial institutions are ‘norm entrepreneurs’ in that they seem eager to engage in Internet cases and take on influential Internet companies. These include the French data protection agency CNIL and the data protection agency of Schleswig-Holstein. Studying this judicial avant-garde may help us predict not only the characteristics of this adjudication but also which judicial and quasi-judicial institutions may take the lead in jurisprudence on algorithmic decision-making and artificial intelligence.
 
We understand courts (and quasi-judicial entities) to be part of a communicative-jurisprudential figuration. We need to further explore their contours and interlinkages. Courts and quasi-courts can be agents of disorder by relying on siloed approach to deciding cases on the Internet. Their narrow view on the cases would suggest that Courts can just partially impact the evolution of digitality. If left unremedied, these tendencies can crystallize into normative fractures that lead to legal uncertainty for states, companies and citizen-consumers.
 
But a different (judicial) world is possible. From the prohibition of the sale of Nazi paraphernalia (as in the early Yahoo cases in France) to the delineation of filtering obligations for intermediaries (as in the Delfi, MTE and Pihl cases by the European Court of Human Rights) and from the development of a fundamental right to the guarantee of the confidentiality and integrity of information technology systems (by the German Federal Constitutional Court in 2008) to the “right to be forgotten” (in Google Spain at the CJEU) or more precisely a “European dereferencing right” (in the Advocate General’s opinion in Google/CNIL): courts have shaped the communicative figuration and normative order of the internet in the past. It is now time to take a critical look at the role that courts and quasi-judicial institutions will have on the evolution of the internet, its players and layers, rules and tools, in the future.
 
We see constellations of actors, including courts and quasi-judicial institutions, to have a special but understudied role in the communicative process of Internet governance. For this reason we will host a workshop series dealing with different angles of judicial power in relation to Internet governance.


Photo: Tingey Injury Law Firm / unsplash