Despite recent internal policy changes, Meta continues to struggle with moderating content related to the Israeli-Palestinian war, especially in Hebrew.
Internal policy guidelines obtained by The Guardian from a former Meta employee highlight a complex process for moderating conflict-related content. However, the documents reveal that Meta lacks a similar process to assess the accuracy of moderation in Hebrew and Arabic content.
The former employee, who remains anonymous, criticized Meta’s policies on censoring hate speech about Palestine as biased. This sentiment is shared by Palestinian advocacy groups.
Front-line workers in the information war surrounding the conflict are hesitant to raise concerns due to fear of repercussions or suspicion of their motives.
The documents, dated this spring, highlight Meta’s difficulties in managing content during the volatile conflict, where language and moderation decisions can have serious consequences.
Criticism of Meta’s disparity in moderating content by language has been ongoing, with whistleblower Frances Haugen raising concerns about the company’s approach.
Source: www.theguardian.com
