gained widespread international attention in 2012. The Muslim community called the Rohingya lived in Rakhine state. They were violated to the extreme and were forced to flee from their homes. I will not go through this conflict in detail but use it in order to explain and discuss how Facebook could be such a big part of this picture and receive the amount of criticism that it has. How come Facebook ended up in the middle of the attention surrounding this particular conflict, and what can it tell us about Facebook's position among local users and my informants? I wish to be clear regarding one crucial fact: Facebook did not cause the Rohingya crises, which I later use as a case, and that fact is vital to acknowledge. They did, however, create a platform where opinions got to flourish and spread, as the Human Rights Council report states: “Facebook has been a useful instrument for those seeking to spread hate, in a context where, for most users, Facebook is the internet” (HRC, 2018, p. 14). It could be easy to give Facebook the blame for allowing the situation to grow into what it developed to be, but there are always several sides to a story. According to the articles in Reuters (Stecklow, 2018b) and Wired (McLaughlin, 2018), Facebook did not take action early enough. They reported that Facebook in 2015 only had 2 employees that could read Burmese and could work with reported content. None of them were stationed in Myanmar. Later on, this work was outsourced to a company situated in Kuala Lumpur, still not with enough Burmese speaking employees to be able to go through reported post fast enough, but things seem to be moving in the right direction (McLaughlin, 2018). However, Facebook is not entirely to blame for the situation. Facebook´s algorithms struggle with reading and interpret Burmese scripts; this is alleged because of font issues. It should come as no surprise then, that Zawgyi is part of the problem. As addressed in chapter 4., the codes in Zawgyi are different from Unicode, and if the algorithms are designed to interpret Unicode, Zawgyi would, once again, become an issue. As far as I can tell, it would suggest that the widely used Zawgyi is the main reasons the algorithms ‘struggle’ within the interpretation and translation. When the algorithms are not able to read and sort out the discriminating content, naturally they would not be able to stop offensive content in such a preferable degree. This technical issue does not just limit the security part where algorithms fail; it is also causing translation problems. Facebook´s translation-tool fails to give an accurate translation of content. Reuter´s example of hateful speech shows how translations are failing. The original Burmese post said: “Kill all the kalars that you see in Myanmar; none of them should 80