Twitter Complaint Hotline Operator

On 8 January 2021, two days after the Capitol Riot, Twitter declared the “permanent suspension” of Donald Trump’s official account, “due to the risk of further incitement of violence”. Elon Musk, Twitter’s conspicuous new owner, famously spoke out against the ban, raising concerns that he could undo it upon gaining control of the social media platform. Mr. Musk was quick to dispel these fears. He first reassured advertisers that he has no plans to relax content moderation standards. In addition, he announced that the reactivation of banned accounts is subject to the creation of a body that will assist Twitter in making such decisions. “Twitter will be forming a content moderation council with widely diverse viewpoints”, he twitted, clarifying that “[n]o major content decisions or account reinstatements will happen before the council convenes”. Anyone feeling a sense of déjà vu?

Begging for a call from Mr. Musk

On reading about a “content moderation council”, it is hard not to think of the Oversight Board that Facebook (now Meta) helped set up and which – with funding comparable to that of the International Court of Justice, an illustrious membership, and well over a million complaints a year – stands as a powerful presence in the field of adjudication about permitted speech on social media. Interestingly, the Oversight Board too wondered how it might relate to the creature that Twitter is hatching: wouldn’t Twitter’s council be just like us and therefore redundant? After such a reckoning, there must have been an exchange of views among the Board members and staff (and perhaps a call with Mr. Zuckerberg), which within less than two hours led to a diplomatic advance in the form of a quote tweet:

Independent oversight of content moderation has a vital role to play in building trust in platforms and ensuring users are treated fairly. This is a model we have been proving since 2020. We would welcome the opportunity to discuss Twitter’s plans in more detail with the company.

So far, little is known about Mr. Musk’s plans, apart from the notion that the council will include “the civil rights community and groups who face hate-fueled violence”. But the unveiling may be just around the corner. On 2 November, he stated that “Twitter will not allow anyone who was de-platformed for violating Twitter rules back on platform until we have a clear process for doing so, which will take at least a few more weeks”. This timetable seems a little optimistic, considering that it took almost three years to get the Oversight Board up and running. Even if Mr. Musk is thinking of a less complex arrangement – perhaps just an ad hoc panel – assembling an authoritative and balanced lineup would be hard in such a short time. The selection would be subject to intense scrutiny by both detractors and supporters of Mr. Trump as well as other prominent political figures. Moreover, whatever one thinks of Meta’s Oversight Board, there is little doubt that it set a high standard for the industry, creating expectations that other players must live up to. A more intriguing possibility would therefore be that Mr. Musk turns the task over to the Oversight Board itself. This scenario seems excluded for now: the above-quoted announcement clearly alludes to a Twitter-made body. However, Mr. Musk is a mercurial type and may well change his mind.

There are a couple of good reasons for accepting the Oversight Board’s invitation. The first is that it would only take a tiny fraction of what Mr. Musk spent to buy Twitter to accede to the Oversight Board’s system and piggyback on the larger investment that Facebook made to set it up. Secondly, if Mr. Musk’s main concern is the fate of Trump’s account, he can easily predict what the Oversight Board’s decision would be. In fact, the Board already settled the similar case of Mr. Trump’s indefinite suspension from Facebook and Instagram, ruling that the sanction was exorbitant and requiring Meta to set a deadline within which to reconsider the matter. In response, the company turned the open-ended sanction into a two-year suspension, expiring on 7 January 2023. This is a decision that Twitter could be tempted to make its own, pending an overhaul of its content moderation system.

But apart from the usual travails over Trump, Mr. Musk’s tweets about a future content moderation council, together with the Oversight Board’s quick reaction, raise questions about the shape such a system could take and the role the Board might play in relation to it. As we shall see, three different scenarios may arise, depending on whether Twitter accedes to the Board’s mechanism, empowers the Board by other means, or sets up its own body. Before discussing these options, it is indispensable to say a few words about the Oversight Board’s unique nature.

Where does the Oversight Board fit in the DSA?

As there is an ongoing debate at Verfassungsblog on the freshly-minted EU Digital Services Act (DSA), it is apt to start by asking in which box of the DSA the Oversight Board falls, if any. One might be tempted to say that the Board is part of Meta’s “internal complaint-handling system” within the meaning of Article 20 of the DSA. But this is not the case. Under Article 20(4) DSA, “[p]roviders of online platforms shall handle complaints submitted through their internal […] system in a timely, non-discriminatory, diligent and non-arbitrary way”. This is what batteries of moderators, mostly recruited by Meta’s contractors, strives to do. The logic of the Board’s operation is in some respects antithetical. The Board takes up just a few exemplary cases – around 1 in 50,000 – and disposes of them with court-like care: it is discriminatory and slow. But this is just a hint that the Board does not fit into a company’s internal remedial system. The conclusive ground is that the Board not only is an entity distinct from Meta, like the contractors in charge of ordinary content moderation, but also one whose rulings, unlike those of the contractors, are not attributable to Meta and are indeed binding on it (Article 4 of the Oversight Board Charter).

Still less could it be argued that the Board engages in “out-of-court dispute settlement” within the meaning of Article 21 of the DSA. To operate in this quality, the Board would have to be “established” in a EU Member State, which it is not. If it were, it would still fail to fulfil several criteria for certification by national Digital Services Coordinators, including that of being “financially independent” of any platform provider (Article 21(3)(a) DSA). Although Meta funds the Board in a manner designed to safeguard independence (see below), the arrangement no doubt falls short of the DSA’s requirements. The Board ultimately fits the residual category of “other available possibilities of redress” of which the platform provider must inform the complainants upon the exhaustion of internal appeals (Article 20(5) DSA). Therefore, recourse to out-of-court settlement, which requires such previous exhaustion, can be made in parallel with the filing of a complaint to the Board. The DSA fails to capture anything but a blurred image of the Board – what is it?

A global institution

On a cursory appraisal, the Oversight Board is easily mistaken for a corporate entity established under US law and entrusted with an atypical task, rather like the ICANN. In reality, the way in which the Board is legally structured is more complex. In 2019, Facebook set up a trust under Delaware law and, in its capacity as the sole settlor, irrevocably deposited USD 130 million to provide for the Board’s operation for an initial period of six years. The trust then created a limited liability company that acquired the facilities and staff to give the Board administrative support and – crucially – entered into service contracts with individual Board members. But it was not the company that selected them. This task was first performed by Facebook in concert with the four (Facebook-appointed) co-chairs of the Board, and now it is up to the Board itself, whose power of self-reproduction realizes an internationalist lawyers’ old dream (see, e.g., Scelle, at 105-107). As Professor Post remarked, “the Board itself doesn’t have legal personality, because it consists of independent contractors, and yet the Board […] is the locus of decision-making power and of independence”.

In a previous post, I argued that the Oversight Board is a legal person under global law. This proposition proceeds from the following assumption: silence about an entity’s legal-system membership – a salient feature of the Board – implies global legal membership if matched by a credible claim to global significance, couched in the language of law. Said silence is hardly surprising if one considers that global law does not possess its own rule of recognition (Walker, at 18-19). It is therefore not a legal system, but it is law nonetheless. In this sense, its situation is analogous to that of international law according to Chapter X of H.L.A. Hart’s The Concept of Law. Alternatively, one could regard the Board as legally non-existent. However, to view it that way would be unrealistic, given that Board is the linchpin of a large organisation, relies on distinguished legal experts, primarily applies international human rights law, and adjudicates with the allure of an international court.

The fact that the Oversight Board inhabits the elusive realm of global law does not mean that it can act free of constraints. For instance, the Board could not enter into an agreement with Twitter regardless of the ties that bind it to Meta. Such constraints are mainly defined by the Oversight Board Charter, an act passed by Mr. Zuckerberg himself, apparently under no legal system – again, a silence that matches global law’s formlessness. The Charter did not establish the Oversight Board. By its own terms, it set “the framework for creating” it. Through the Charter, Mr. Zuckerberg empowered a group of people to constitute the Board as something other than a legal entity under US law. And since global law, rudimentary as it still is, offers no corporate law toolbox, such “chartered community” created the Board through practice, making it a custom-based institution. The Charter, as a piece of global law, regulates inter alia the relationship between the Board, Facebook (now Meta) and the above-mentioned trust. It therefore refers to the relevant US (Delaware) law – the law under which the trust is established – making it applicable in the field of global law. One must take these circumstances into account to see how Twitter could establish a relationship with the Board, should Mr. Musk wish to.

Mr. Musk’s options

As mentioned above, Twitter has at least three available options. The first consists in joining the Oversight Board’s mechanism. The Trust Agreement (Sections 1.2 and 2.1) allows for the accession of other platform providers on condition that they are “US Person[s]”, contribute additional property to the trust, thus sharing in the considerable financial burden of maintaining the structure, and commit to abide by the Board’s decisions in accordance with the Charter. Since such an accession would require the amendment of the Trust Agreement and the Charter, it could not take place without the consent of Meta, a majority of the Individual Trustees, as well as the Board itself deciding by absolute majority (Section 1(5) of the Trust Agreement; Article 6(1) of the Charter). Alternatively, Twitter could set up its own trust and demand that it become a member of the limited liability company alongside the original trust, which is currently the company’s sole member (Section 3.2 of the LLC Agreement). The consent of all the parties, including the Board, would be required in this case as well, again because Twitter’s accession would require the amendment of the Charter.

By contrast, a reworking of the rules governing content moderation on the respective platforms would be unnecessary, as the Board already applies different sets of rules depending on whether it deals with Facebook or Instagram cases. However, the assessment of such regulations against international human rights law – something the Board routinely does – would likely lead to their convergence on key issues. Twitter’s accession might bring about an increase in the Board membership and perhaps the creation of a dedicated panel, a step that would not affect the consistency of the case law as panel decisions are subject to review by the full Board (Article 1, Section 3.1.8 of the Oversight Board Bylaws).

If the above feels to Mr. Musk like turning Twitter into Meta’s vassal, he could empower the Board following a purely global law route. He could do it by replicating Mr. Zuckerberg’s moves, namely, issue a charter vesting power in the Board, set up parallel trust and limited liability company, and transfer to those structures the funds needed to finance the Board’s additional operations. Meta’s consent would not be required. It is true that the Charter, at Article 1(4), stipulates that the Board “will have no authority or powers beyond those expressly defined by this charter”. But Musk’s would be another charter, granting the Board additional powers under global law – provided the Board accepts such a grant. The Board’s consent would of course be required and would mark the conclusion of an agreement under global law, since the Board, which lacks legal personality under US law, cannot contract under that law.

The third and final option would be for Twitter to fashion an institution comparable to the Board. Such a body would likely take into consideration the Board’s case law but could also generate a distinct one, thereby offering its counterpart some dialectical stimulation. Of course, the foregoing assumes that Mr. Musk prefers the elite art of adjudication to a users’ plebiscite – “vox populi vox dei”, he recently posted – and to his own will, channeled by a panel of like-minded people. After all, Mr. Musk’s official Twitter account introduces him as the “Twitter Complaint Hotline Operator”.