Crime as Cognitive Constraint: Facebook's Role in Myanmar's Incitement Landscape and the Promise of International Tort Liability.

• Author: Domino, Jenny
• Position: Atrocity Prevention: The Role of International Law and Justice

Table of Contents Table of Contents Introduction I. Background II. Gap A. Prohibited speech under international law 1. International human rights law 2. International criminal law B. Legal responsibility for direct and public incitement to commit genocide 1. State responsibility 2. Individual (criminal) responsibility C. Corporate responsibility III. Critique A. Corporate criminal liability B. Crime as cognitive constraint: Facebook as a "useful instrument" for hate speech in Myanmar IV. The Promise of Tort A. Rationale 1. Sanction and criminal law's design B. Cause of action 1. To whom the duty applies: the public-private flip 2. Duty 3. To whom the duty is owed; causation C. Design V. Conclusion Introduction

Is Facebook's role in the spread of incitement in Myanmar criminal? In 2018, the United Nations Independent International Fact-Finding Mission on Myanmar (FFM) described Facebook's "significant role" in the spread of incitement to discrimination and violence against Myanmar's Rohingya Muslims. (2) Although the FFM described in detail the speakers' responsibility for the Facebook posts, the precise nature of Facebook's responsibility in moderating (or failing to moderate) harmful content was unclear.

Facebook's tragedy in Myanmar is a striking example of how international law's current regimes of responsibility for incitement to genocide results in a gap. The extant regime, with its focus on states and natural persons, is incapable of fully accounting for the different actors and technologies involved in the rich and complex narrative of mass violence. One popular response to the lack of corporate liability is to expand domestic and international prosecution of legal or juridical persons. (3) However, by evaluating the role of social media platforms in the spread of incitement, particularly the damage wrought by Facebook in Myanmar, this article hopes to contribute to that literature by showing how crime can operate as a cognitive constraint in appreciating different modalities of corporate involvement in situations of mass atrocity. By emancipating ourselves from this dominant cognitive frame, we can imagine a broader network of international liabilities. One possibility is tort. For this purpose, Facebook's ban of Myanmar's commander-in-chief from the platform (4) merits further consideration. It remarkably embodies a regulatory paradox, where the traditional roles of regulator (state) and regulated entity (non-state) are reversed. This article examines the implications of the exercise of this awesome power in articulating an international tort liability and establishing a related mechanism.

At an international symposium held in 2019 on the subject of Law and Atrocity Prevention, I was asked to speak on a panel about "regulating social media that fosters atrocity crimes." (5) The word choice for the panel discussion--"fosters"--was very telling. It reveals the uncertain legal character of social media companies' involvement in instances of incitement to violence posted on the platform, including the crime of direct and public incitement to commit genocide under the Genocide Convention and the Rome Statute (alternatively, "incitement to genocide"). (6) Such ambivalence was similarly apparent in the FFM report. The FFM found that ultranationalist monks, military leaders, and government officials were responsible for posts that constituted incitement to violence and persecution against the Rohingya Muslims of Myanmar's Rakhine State. (7) Discriminatory rhetoric and calls for violence created a social environment which facilitated the armed confrontations between Myanmar's military--the Tatmadaw--and the non-state armed group, Arakan Rohingya Salvation Army. This, in turn, resulted in the present Rohingya refugee crisis. (8)

In contrast, the FFM's description of Facebook's involvement was less straightforward:

The Mission has no doubt that the prevalence of hate speech in Myanmar significantly contributed to increased tension and a climate in which individuals and groups may become more receptive to incitement and calls for violence. This also applies to hate speech on Facebook. The extent to which the spread of messages and rumours on Facebook has increased discrimination and violence in Myanmar must be independently and thoroughly researched, so that appropriate lessons can be drawn and similar scenarios prevented. Similarly, the impact of the recent measures taken by Facebook to prevent and remedy the abuse of its platform needs to be assessed. (9) The FFM found that social media played a "significant" role in the spread of incitement in Myanmar, and that Facebook served as a "useful instrument" for hate speech (10) "in a context where for most users Facebook is the Internet." (11) To be used, to serve as an instrument--by using the passive voice to describe Facebook's actions, the FFM obscured the platform's agency and avoided the attribution of legal responsibility to Facebook as a company engaged in the business of content moderation. Such silence is to be expected. International law has only recognized two regimes of legal responsibility--for states and natural persons--in regulating incitement to genocide. The Genocide Convention imposes on states the duty to prevent and punish genocide within their borders (including incitement to genocide), (12) while international criminal law prosecutes natural persons for statements amounting to direct and public incitement to commit genocide. (13) In contrast, the UN Guiding Principles on Business and Human Rights (UNGPs) provide a template to guide corporate conduct without assuming the form of legal obligation. (14)

Calls for accountability have been consistent with this framework. The FFM recommended investigation for possible genocide, crimes against humanity, and war crimes; the UN Independent Investigative Mechanism for Myanmar was established soon after to collect and preserve evidence that may be used in a court of law. (15) Separately, the Prosecutor of the International Criminal Court (ICC) is now looking into acts of deportation that spilled onto neighboring Bangladesh, a state party to the Rome Statute. (16) This would circumvent Myanmar's status as a non-state party to the treaty and allow the prosecution of individuals found to be most responsible for the deportation of the Rohingya. (17) Demand for state responsibility ensued. In November 2019, The Gambia filed an application before the International Court of Justice (ICJ) concerning Myanmar's alleged violation of its erga omnes obligations under the Genocide Convention, which Myanmar is a party to. (18) Importantly, The Gambia devoted a substantial portion of its application to describe the "hate propaganda" against the Rohingya. (19) In an unprecedented order, the ICJ granted jurisdiction to rule on The Gambia's request for provisional measures and enjoined Myanmar to prevent further acts of genocide within its borders, including the crime of direct and public incitement to commit genocide. (20)

Myanmar's incitement landscape thus presents a striking picture of how international law's current regimes of responsibility appear to result in a gap, incapable of fully accounting for the different actors and technologies involved in the rich and complex narrative of mass violence. Recent trends to address the lacuna for corporate liability in mass atrocity contexts involve hardening the soft law framework contained in the UNGPs through a legally binding instrument, which would impose on states the obligation to prosecute juridical persons at the national level, (21) and to expand the coverage of the Rome Statute to allow the prosecution of juridical persons before the ICC. (22) These movements comprise two sides of the same coin. Although my paper is not focused on individually evaluating these approaches, much less urge proponents to abandon criminal law altogether, I argue broadly that such approaches exhibit an unquestioning attitude towards the primacy of crime to conceptualize harm, which can operate as a cognitive constraint in appreciating different modalities of corporate involvement in mass atrocity contexts.

As various scholars, legislators, and policymakers develop theoretical and policy approaches to regulate social media, (23) this article hopes to contribute to the conversation by showing how the language of crime can limit our conceptual thinking of harm. By freeing ourselves from this dominant cognitive frame, a broader network of international liabilities is up for imagining. One possibility is international tort liability. For this purpose, Facebook's ban of Myanmar's commander-in-chief from the platform deserves consideration. (24) It remarkably embodies a regulatory paradox, where the traditional roles of regulator (state) and regulated entity (non-state) are reversed. This highlights the agency of social media platforms and the process of content moderation that lies at the heart of their business. As Tarleton Gillespie defines, a "platform" is an online site or service where content is provided by users but the company offering the technology moderates user content and activity as an "essential" (rather than "ancillary") undertaking. (25) This article examines the exercise of private regulation of state actors' speech--in the form of content moderation--in distilling generalizable principles for an international tort liability and the institutional design of a related mechanism, extending Maya Steinitz's blueprint for an International Court of Civil Justice (alternatively, ICCJ). (26)

It bears emphasizing that the focus on Facebook is not intended to single out one platform; rather, it is meant to provide a legal-theoretical response to what other experts have already commented upon. (27) It is also meant to function as a cautionary tale for technologies operating in incitement contexts. (28) The term "speech" used in this article refers to the full range of...