4/10/22, 10:45 Oversight Board | Independent Judgment. Transparency. Legitimacy. expression are imposed by a state, they must meet the requirements of legality, legitimate aim, and necessity and proportionality. The Board applies these international standards to assess whether Meta complied with its human rights responsibilities. I. Legality (clarity and accessibility of the rules) The principle of legality requires laws that States use to limit expression to be clear and accessible, so people understand what is permitted and what is not. Further, it requires laws restricting expression to be specific, to ensure that those charged with their enforcement are not given unfettered discretion (General Comment 34, para. 25). The Human Rights Committee has warned that “offences of ‘praising’, ‘glorifying’, or ‘justifying’ terrorism, should be clearly defined to ensure that they do not lead to unnecessary or disproportionate interference with freedom of expression. Excessive restrictions on access to information must also be avoided” (General Comment No. 34, at para. 46; see also: UN Special Rapporteur on counter-terrorism and human rights at paras 36-37(Report A/HRC/40/52)). Following its approach in previous cases, the Board applies these principles to Meta’s content rules. While the Board welcomes that Meta’s policies on Dangerous Individuals and Organizations contain more detail now than when the Board issued its first recommendations in this area, serious concerns remain. For users reporting on the Taliban, it is unclear whether the Taliban remains a designated dangerous entity when it forcibly removed the recognized government of Afghanistan. The Board has previously recommended that Meta discloses either a full list of designated entities, or an illustrative one, to bring users clarity ("Nazi quote" case, "Ocalan’s isolation" case). The Board regrets the lack of progress on this recommendation, and notes that while the company has not disclosed this information proactively, whistleblowers and journalists have sought to inform the public by disclosing a version of the “secret” list publicly. As noted previously in this decision (see section 8.1), the definition of “praise” in the public-facing Community Standards as “speaking positively about” a designated entity is too broad. For people engaged in news reporting, it is unclear how this rule relates to the reporting allowance built into the same policy. According to Meta, this allowance permits news reporting even where a user praises the designated entity in the same post. The Board finds the relationship between the “reporting” allowance in the Dangerous Individuals and Organizations policy and the overarching newsworthiness allowance remains unclear to users. In the “Shared Al Jazeera post" case, the Board recommended that Meta provides criteria and illustrative examples in the Community Standards on what constitutes news reporting. Meta responded in the Q1 2022 implementation report that it was currently consulting with several teams internally to develop criteria to help users understand what constitutes news reporting. It said it expects to conclude this process by Q4 2022. The Board remains concerned that changes to the relevant Community Standard are not translated into all available languages and there are inconsistencies across languages. Following the Board’s “Shared Al Jazeera post” decision, the US English version of the Dangerous Individuals and Organizations policy was amended in December 2021 to change the discretionary “we may remove content” to “we default to remove content” when a user’s intentions were unclear. However, other language versions of the Community Standards, including in Urdu and UK English, do not reflect this change. While the company has stated publicly in response to previous recommendations from the Board ( Meta Q4 2021 Quarterly Update on the Oversight Board) that it aims to complete translations into all available languages in four to six weeks, it appears the relevant policy line for this case has not been completed after five months. Therefore, the policy is not equally accessible to all users, making it difficult for them to understand what is permitted and what is not. The Board is also concerned that Meta has not done enough to clarify to its users how the strikes system works. While a page on “Restricting Accounts” in Meta’s Transparency Centre contains some detail, it does not comprehensively list the feature-limits the company may apply and their duration. Nor does it list the “set periods of time” for severe strikes as it does for standard strikes. This is especially concerning because severe strikes carry more significant penalties and there is no mechanism in place for appealing account-level sanctions separately from the content decision. Even when the content is restored, feature-limits cannot always be fully reversed. In this case, for instance, the user had already experienced several days of feature-limits which were not fully rectified when Meta reversed its decision. That two Urdu-speakers assessed this content as violating further indicates that the praise prohibition, and its relationship to the reporting allowances, is unclear to those tasked with enforcing the rules. Content reviewers are provided with internal guidance on how to interpret the rules (the Known Questions and the Implementation Standards). The Known Questions document defines praise as content that “makes people think more positively about” a designated group. This is arguably broader than the public-facing definition in the Community Standards, making the meaning of “praise” less about the intent of the speaker than the effects on the audience. Also, neither the Community Standards, nor the Known Questions document constrain the reviewer’s discretion on restricting freedom of speech. Standard dictionary definitions of “praise” are not this broad, and as phrased the rule captures statements of fact, including impartial journalistic statements, as ll i i I t th B d’ ti https://www.oversightboard.com/decision/FB-U2HHA647 M t l ifi d th t th ti ll ll d t l j li t t 9/12