Ongoing violations by the military and military-linked accounts and Pages since the February 1 coup, including efforts to reconstitute networks of Coordinated Inauthentic Behavior that we previously removed, and content that violates our violence and incitement and coordinating harm policies, which we removed. The coup greatly increases the danger posed by the behaviors above, and the likelihood that online threats could lead to offline harm. Factors 1 and 2 were true long before the coup, while Factors 3 and 4 were true in the first week of the coup as much as in the third week. This is also not the first time that the Tatmadaw disregarded the people’s vote. Further, rumors of a coup spiked in late Jan. this year, prompting diplomatic missions in Myanmar to release a joint statement urging the military to recognize the election results. Facebook’s responses had also been partially preempted: By Feb. 24, Facebook and other social media platforms were already banned in Myanmar as part of the military junta’s series of network disruptions which, since the coup on Feb. 1, has involved internet and mobile network shutdowns and social media and website bans. Despite the local social media ban, however, Facebook’s decision still carried weight as many people within Myanmar continue to access the platform through virtual private networks (VPN). Overall, Facebook’s response particularly pales in comparison to its relatively swift action to de-platform former President Trump soon after the U.S. Capitol riots and then refer the matter to the Facebook Oversight Board. Although abhorrent, the violence at the U.S. Capitol was mild compared to the scale of violence called for and facilitated by military-linked Facebook accounts in Myanmar. And the threat to democracy posed by the Jan. 6 insurrection was dwarfed by the actual overthrow of democracy on Feb. 1 in Myanmar and the international crimes several years earlier during the Rohingya crisis. These contrasts reveal a broader problem with Facebook’s approach to content moderation in the most fragile contexts. A Global Conversation Centered On At-Risk Populations Facebook’s inconsistent and often-belated de-platforming approach in Myanmar should invite deeper reflection on the parameters of social media access provided to world leaders of illiberal and authoritarian regimes. In its decision on the Trump ban, the Facebook Oversight Board made a policy recommendation to Facebook to “publicly explain” the applicable rules when imposing account-level sanctions against influential users, including its strikes and penalties process. Facebook should take up this recommendation and clarify how it enforces such policies abroad. This problem is also not unique to Facebook. Other platforms such as TikTok and YouTube have respectively moderated Tatmadaw soldiers and video channels for violent content, but have been vague about these content decisions. Further, beyond formally committing to provide access to remedy in line with the UNGPs, social media companies should explore how various forms of remedy and reparation (including compensation, rehabilitation, and satisfaction in the form of public apologies, memorials, and truth-telling) ought to be made available to communities in Myanmar affected by the adverse human rights impacts that their technology or business operations have engendered. As suggested by Rohingya refugees in Bangladesh, this would include engaging with human rights victims, responding clearly and promptly to requests, providing free internet access to refugee camps, and using their influence to promote an open internet, especially in the region where majority of their users are located.