Over the past years, Facebook as well as other online platforms faced constant criticism and pressure from civil society, lawmakers, and governments regarding their role in content moderation. As a response, Facebook for the last two years sought a way to handle the decision-making process in a more transparent, accountable, and fair way, while at the same time diverting responsibility from itself and its CEO and sole controller, Mark Zuckerberg. Now, it finally rolls out its plan for an independent Oversight Board, also referred by some, including Mark Zuckerberg, as a “Supreme Court”. Major questions arise as to its mandate, the establishment of global free speech standards and the reactions of national courts.
The New Oversight System
This week, Facebook released the final “Oversight Board Charter”, together with an updated version of its community standards. The Charter describes a review system entailing three institutional players, two of them to be newly established.
The Oversight Board will be initially comprised of 11, later up to 40 members. It will review content decisions on all of Facebook’s platforms. This entails decisions to “allow or remove” content as well as the so-called designations of content, like warning screens or potential classifications as fake/untruthful. Affected users and Facebook can appeal to the Board, though it will have discretion to pick cases based on “significance and difficulty”. Its judgements “will be binding on Facebook”. In addition, the Board will be able to issue case-specific “policy advisory statement(s)” while Facebook will have the right to ask for general “Policy Guidance”.
A Trust established and funded by Facebook will secure the Board’s governance, compensating the Board’s members and providing operational support. More importantly, it will also appoint new members upon recommendation by the Board and will have the right to remove members for breaches of a (to be established) code of conduct, though not for any content decision.
Facebook will appoint the Board’s initial two members. These will then, together with Facebook, suggest the other members in accordance with the established Selection Process. Most importantly, Facebook will be implementing the Board’s decisions – thus letting the Board “overrule Zuckerberg”. However, it will not do so in case it “could violate the law”. The Board’s guidance yet is merely advisory and only to be “consider(ed)”. Facebook will support the Board’s decision-making “to the extent that requests are technically and operationally feasible and consistent with a reasonable allocation of Facebook’s resources”.
Questions Regarding the Board’s Mandate
The Board in fact does not resemble an all-encompassing Supreme Court. Rather, the Charter tries to confine its mandate to “review and decide on content”. This includes whether “content should be allowed or removed completely” and “additional questions related to the treatment of content”, such as warnings or clarifications. The Board thus decides on the limits of free speech on the platform. However, it seems unlikely that the Board will be able to limit its decisions to an assessment of free speech. Rather, the board will ultimately not be able to avoid expanding its mandate to many other important fields.
- As part of the review process it will be unavoidable for the Board to decide on questions of judicial rights and fair process regarding both its own and Facebook’s handling of content decisions. It will be faced with answering whether users have “exhausted appeals”, the extent of the right to “submit relevant and informed written statements”, and standing of groups considered as “immediately depicted or impacted”. That alone is enough to open a Pandora’s box of procedural problems. On top of that, the Board will have to position itself towards issues such as statutes of limitations, necessary substantiation, legal assistance, equality of arms, and access to justice in general.
- The Board will have to take a stand on if and how far it can go against automated content review. It certainly can review automated content decisions in every single case. However, it seems less clear if it could address systematic problems with automated decision making. Certainly, it has some tools at its disposal, as it can ask Facebook to provide information that is “reasonably required for the Board to make a decision”. Nevertheless, Facebook can reject such a request in case it is not “technically and operationally feasible”. Explaining and handing over code could cross that line.
- It is unclear how far the Board is allowed to delve into questions of policing, criminal law and sanctioning. When Facebook blocks users for some time or deletes their accounts, it goes beyond deciding on the permissibility of content. Rather, it takes measures of preventive policing or ex-post punitive justice. Such decisions will certainly be appealed before the Board – even if it remains unclear if they are covered by its mandate.
On the other hand, Facebook aims at excluding some pressing issues from the Board’s mandate, though it would have made sense to be included. First, the matter of election meddling and political ads. It will be important to see if the Board picks them up anyway. One potential hook is, that even if they are not content in a narrow sense, they still relate to the broader question of what can and cannot be on the platform. Second, the ‘Facebook algorithm’ itself. The fact that the Board will not be able to hear cases on the ranking and dissemination of content constitutes an important caveat. It is this algorithm which has been blamed for leading to the overspreading of extremist content (also here) or even its creation. This will cause practical difficulties in the Boards work: so-called “shadowbanning” makes it hard to draw the line between deleting and just downranking.
Establishing a Global Free Speech Standard
The Board’s members will face the unique and seemingly impossible task of establishing a global standardfor permissible online speech. Although, as others have acknowledged, there often is no “right” answer, the Board nevertheless must strive to find ONE globally acceptable solution. This endeavor might take the establishment of transnational private law to a new level. The sought-after “diversity” of the members, albeit necessary for the Board’s legitimacy, will make the task particularly demanding.
The Charter and the updated community standards figure as Facebooks ‘Bill of Rights’. Putting forward a clear commitment to “voice” as the governing principle and free speech as the primary right (calling it a “fundamental human right”), Facebook has arguably already made one of the most “difficult choices” itself. However, both documents acknowledge that free speech “must be balanced against” other values. Facebook seems to be leaning towards a Europe-styled balancing approach, though severe fights over this direction among the Board’s members seem inevitable. This could be aggravated by the fact that when deciding on the limits of free speech, the Board members will have to also assess the value and importance of the counterbalancing human rights. The Charter itself names “authenticity, safety, privacy, and dignity”. Others will be surfacing, too, like the protection of children and minorities. On top of this, in case of conflict the hierarchy between the two documents is not fully clear – as the Charter is changeable only in case the Board, Trust, and Facebook agree, it seems to reign supreme.
The question of interpretation and hierarchy becomes even more intricate, as the Charter and the community standards incorporate “human rights norms protecting free expression” and “international human rights standards” respectively (in line with the UN Special Rapporteurs suggestions). Thus, the Board will likely try to back up its decisions with (inter-)national human rights jurisprudence (as e.g. the CAS does already today). The selection and weighing among the multitude of potential sources (international human rights bodies, international courts, national courts, etc.) appears to be an endeavor almost bound to fail.
Finally, both the Board and Facebook will have to find ways to engage with the limits on freedom of speech set by national laws. The Charter does not seem to provide much help in this regard. On the one hand, it is going out of its way to assure it respects “the law”: it explicitly says that nothing within it “shall be interpreted in a manner that would result in a violation of law”. On the other hand, however, it also mandates that the “Board will not purport to enforce local law”, making it clear that the Board cannot justify its decisions by reference to local standards. How the line between unitization and fragmentation of free speech on the platform should be drawn remains Facebook’s secret – as well as a major headache for all the actors involved.
The Involvement of National Courts
One of the most important questions, at least from the users’ perspective, will be how courts deal with the procedure established in the Charter. This question could be answered differently in different judicial systems.
From a European and German perspective for example, the first question seems to be if the courts will accept the Board as an arbitral tribunal. This would translate into closing the doors to state courts while at the same time allowing for a limited application of fair trial guarantees (see latest ECtHR, Mutu and Pechstein v. Switzerland, at 96 on CAS). This, however, appears rather unlikely at least as far as German law concerned (I have laid out a roadmap elsewhere), given that the members would be lacking sufficient independence from both parties (see the German FCJ on CAS). Things might be different though under the ECtHR’s jurisprudence (Mutu and Pechstein v. Switzerland, at 139 ff., which has accepted the general setup of CAS).
Even if the Board wasn’t an arbitral tribunal, its establishment could make it harder for users to find redress in national courts. First, courts might demand the exhaustion of Facebook’s remedies. This seems very likely, as already today users arguably have to appeal a content decision before they can take their case to court (see the jurisprudence of the German courts). Second, courts might take the seemingly international nature of the Board’s decisions into account when deciding how much room Facebook should have to decide on the limits of permissible speech. Less as a question of law (although it could be seen as one) than rather one of judicial behavior, judges’ likeliness to scrutinize the Board’s decisions might be diminished. The current strict line of German courts holding Facebook accountable to constitutional standards of free speech could be softened.
Difficulties could also arise in cases demanding expedite proceedings. Although the Charter establishes an “Expedite Review”, it is unclear how much time it will actually take. There might be urgent situations in which national courts feel the need to not await the Board’s decision but take up the case right away. Lastly, it might also be difficult for users to determine whom to sue – Facebook, the Trust LLC, or ultimately both.
Will It Work?
Zuckerberg was not comfortable making “important decisions about speech” – which governments all over the world have effectively eschewed themselves. This is particularly true in the U.S. where – arguably – the U.S. Constitution would prevent the legislator from narrowing the limits of free speech online. In Germany the task of deciding what stays up or should be taken down was specifically placed on Facebook (and other platforms) via the Network Enforcement Act. Before this backdrop, handing this task over to an independent Board suddenly became an attractive option for Facebook.
In fulfilling its task, the Board will not give the “right” answers to the question of how to regulate online speech, but provide legitimacy to its choices in ‘hard cases’. Given its purported independence, plurality, and fair trial guarantees, the Board will, unlike Facebook, enjoy the legitimacy held by state courts at least to a certain degree (see report on Facebook’s transparency). By explaining the reasons behind its judgements, it will enable an ongoing legitimizing academic, “legal”, and public discourse over free speech online. However, there are unavoidable complications and downsides, as detailed above. Most importantly, the Board will still have to prove its independence from Facebook while it tries to establish its global reputation. If it succeeds, it could become – for better or worse – a new “model for our industry”.
The job of the Board will not necessarily make it unpopular. Quite the contrary, chances are the Board will be seen as the guardian of a safe, open, and principled digital public space – the same way in which many Constitutional and Supreme Courts around the world are regarded in terms of analog spaces. In case you are interested: the Board is now hiring.