2020 Statelessness & Citizenship Review 2(1) pledged to deploy dozens of Burmese-language moderators to ensure that Facebook would not be used for similar purposes in the future. 75 A recent report entitled An Independent Assessment of the Human Rights Impact of Facebook in Myanmar 76 looked into Facebook’s responsibility in spreading online hate and inciting offline violence against the Rohingyas in Myanmar concluding that Facebook was not doing enough to help prevent its platform from being used to foment division and incite offline violence, agreeing that Facebook can and should do more to prevent such acts, and stated having had ‘invested heavily in people, technology and partners to address the abuse of Facebook in Myanmar’. 77 The report also set out some recommendations to advise Facebook on mitigating the adverse human rights impact, while maximising the opportunities for freedom of expression, digital literacy, and economic development. 78 In order to follow up on BSR’s recommendations, Facebook has put in place a team with native Myanmar language speakers who have a better understanding of the country context to work on issues specific to Myanmar, reviewing content, as well as ‘improving the development and enforcement of [their] policies’. 79 Additionally, Facebook has undertaken to understand how content that does not otherwise breach their internal guidelines may have the potential to incite offline violence, alongside an update to its credible violence policy for the removal of misinformation that contributes to violence. On an operational level, Facebook sought to improve their detection capacity by engaging in more proactive detection of hate speech in Myanmar, and taking an aggressive stance on the deletion of accounts that mislead others about who they represent to be. 80 As part of these steps towards the identification of violence, Facebook also invested in the extended use of artificial intelligence, allowing their software to engage posts that contain ‘graphic violence and comments that are violent and dehumanising’ while reducing their distribution as they are investigated by the Facebook Community Operations team, who verify whether the given content violates Facebook policies and remove that content to limit its visibility. 81 Four months after Facebook CEO Mark Zuckerberg’s commitment to address content moderation expressed in front of the US Senate, Reuters still found more than ‘1,000 examples of Facebook posts, comments and pornographic images’ regarding the Rohingya and other Muslims calling them such names as ‘dogs’, 75 76 77 78 79 80 81 Adam Taylor, ‘The Big Questions for Marc Zuckerberg on Facebook’s Role in Burma’, Washington Post (online, 10 April 2018) <https://www.washingtonpost.com/news/worldviews/wp/2018/04/10/the-big-questions-formark-zuckerberg-on-facebooks-role-in-burma/>; Anthony Kuhn, ‘Concern Is High Over Facebook’s Role in Fanning Violence in Myanmar’, NPR (online, 14 June 2018) <https://www.npr.org/2018/06/14/619854631/concern-is-high-over-facebooks-role-infanning-violence-in-myanmar?t=1591748452826>. Alex Warofka, An Independent Assessment of the Human Rights Impact of Facebook in Myanmar (Report, Facebook 5 November 2018) <https://about.fb.com/news/2018/11/myanmar-hria/>. The assessment — according to the report — ‘was completed by Business for Social Responsibility — an independent non-profit organisation with expertise in human rights practices and policies’. ibid. ibid: (1) building on existing governance and accountability structures, (2) improving enforcement of content policies, (3) increasing engagement with local stakeholders, (4) advocating for regulatory reform and (5) preparing for the future. ibid. ibid. ibid. 18