4/10/22, 10:45 to verify the requested data in the time available. Oversight Board | Independent Judgment. Transparency. Legitimacy. 7. Public comments The Oversight Board received and considered six public comments related to this case. One of the comments was submitted from Asia Pacific and Oceania, four were from Europe, and one was from the United States and Canada. The submissions covered the importance of access to social media to people who live in or near Afghanistan, concerns about limitations on the discussion of designated groups, and the public interest in allowing a wider range of media reporting on the Taliban’s actions. Several public comments argued that the reliance of Meta's Dangerous Individuals and Organizations policy on the vague terms praise and support may suppress critical political discussion and disproportionately affect minority communities and the Global South. Public comments also criticized Meta for using US law as an “excuse” to prohibit “praise” of designated groups, rather than be transparent that it is Meta’s policy choice to restrict more expression than US law requires. To read public comments submitted for this case, please click here. 8. Oversight Board analysis This case is significant because it shows that a lack of clarity in the definition of praise seems to be resulting in uncertainty among reviewers and users. It also considers important issues of content moderation as it applies to gender discrimination and conflict. This case is difficult because there is an interest in ensuring that terrorist groups or their supporters do not use platforms for propaganda and recruitment efforts. However, this interest, when applied too broadly can lead to censorship of any content that reports on these groups. The Board looked at the question of whether this content should be restored, and the broader implications for Meta’s approach to content moderation, through three lenses: Meta's content policies, the company’s values and its human rights responsibilities. 8.1 Compliance with Meta’s content policies I. Content rules The Board finds that the content falls within the allowance that “users may share content that includes references to designated dangerous organizations (…) to report on (…) them or their activities” and is therefore not violating. The content is not violating despite the broad definition of praise provided in the public facing Community Standards, even as the content can be understood as speaking positively about an action of a designated entity, the Taliban, and may give them “a sense of achievement.” The Board notes the Known Questions guidance provided to moderators for interpreting praise is even broader. It instructs reviewers to remove content for praise if it “makes people think more positively about” a designated group, making the meaning of “praise” less about the intent of the speaker but the effects on the audience. Reporting on a designated entity’s claimed intentions to allow women and girls access to education, however dubious those claims, would arguably make people think more positively about that group. Given the instructions they were provided, the Board understands why two human reviewers (in error) would interpret the content as praise. The Board accepts that Meta intends for its Community Standard on Dangerous Individuals and Organizations to make room for reporting on entities Meta has designated as dangerous, even if that reporting also meets the company’s definition of praise. However, the Board does not think the language of the Community Standard or the Known Questions makes that definition clear. As a matter of fact, without specification, “praise” remains overbroad. The Community Standards do not provide any examples of what would constitute acceptable reporting. There is also no internal guidance to moderators on how to interpret this allowance. II. Enforcement action The Board notes that the moderation action in this case occurred after a user began to report the content but never finished the report. An automated system is triggered when this process is initiated, even if the user does not end up completing the report, and so the content was enqueued for human review. The reporting user was not made aware that their action could trigger consequences even if they decide against finishing the report, whereas a user would be told of the consequences of their report if they submitted it. Meta argues in its responses that the “automated report is not tied to the [reporting] user” but the Board finds this noteworthy when the whole process for this case began with a user initiating a report. Also, the Board is concerned that the reporting button does not provide users with sufficient information on the consequences of clicking it. Th f i k i hi (h https://www.oversightboard.com/decision/FB-U2HHA647 l ik df li i i ) h ld h b i d h 7/12