  in response to a religious leader’s post containing allegations that a Muslim man had raped a Buddhist woman. • Another call to kill Muslims was a comment on a widely shared news article, believed to be fake, that a Burmese soldier had been killed by a Muslim. Most of the examples of posts by normal users had few followers or reactions and were not shared widely. However, the most popular post of all the examples in the study was a normal user sharing the alleged restrictions the country of Japan places on Muslims entering and living in the country, which is untrue. This was shared over 18,000 times. Even where the user was not a well-known figure, content relating to current events in Myanmar received the most reaction, such as the Presidential revocation of ‘white cards’, 307 a temporary identification card, from displaced and stateless Muslims applying for citizenship, or advising women to be wary of Muslims during Thingyan (the Buddhist Water Festival in April). Conclusion of Field Research Observations The observation provoking the most serious concern from this short monitoring study is the impact of people in positions of influence, such as politicians or religious leaders, making statements that may incite violence, hatred, or discrimination. These public statements appear to encourage other users to repeat the sentiments, and even go further, such as issuing calls to kill people. This is particularly worrying as Myanmar approaches elections, because they have the potential to incite violence. C. Hate Speech: Recommendations for ICT Companies   307 Identify the potential impacts a company may have: For example, decisions taken by ICT companies on how to tackle hate speech have the potential to impact the right to freedom of expression by: • Providing access to platforms that allow user-generated hate speech content to flourish; • Making their own internal decisions to remove content • Responding to government requests to block access to certain websites or remove particular content that may be hate speech or may be other types of permitted speech that the government has chosen to label as hate speech. Understand the legal framework: As outlined above and in Chapter 4.1 on Freedom of Expression, the legal framework that could be applied to online communications contains vague and undefined terms. While these vague terms could be used to block access to national, racial or religious hatred in line with Article 20 of the ICCPR, those same provisions are so broadly worded that they could result in legitimate content being removed or blocked as well. The Government or other groups’ (such as religious or ethnic groups) may request or require that companies restrict freedom of expression that does not fall within the permitted restrictions under Article 19 or the prohibitions under Article 20. In such cases, an ICT company will find it challenging to Radio Free Asia, Myanmat Authorities Step Up Collection of Temporary Identification Cards (6 April 2015). CHAPTER 4.2: HATE SPEECH 147 4 4.2