Report on Meta's conduct at the Guardian of the Walls: "She did not do enough to deal with inflammatory and offensive content" | Israel Hayom

The technology giant Meta published today (Thursday) a report examining the impact of its policy during Operation Wall Guardian - with an emphasis on monitoring inflammatory and offensive discourse, which led to the company's enforcement, which was reflected in the removal of content, blocking of accounts and cancellation of hashtags.

The report was commissioned by Meta and compiled by Business for Social Responsibility, an organization for human rights, which examined the discourse on Meta's many platforms: Facebook, WhatsApp and Instagram during the Wall Guard operation, which lasted throughout the month of May last year.

Riots during Operation Wall Guard (archive), photo: Reuters

One of the claims in the report states that Meta's actions "had a negative impact on the human rights of Palestinian and Arabic-speaking users." It is also claimed that Meta over-enforced (removal of content) content in the Arabic language: "The rate of proactive monitoring of content In the Arabic language that may be found in violation of Meta's policy terms - higher than the monitoring rate of Hebrew content.

Content in Hebrew was subject to lack of enforcement."

The report attributes one of the reasons for this to the fact that "Meta policy combines legal obligations related to terrorist organizations and the fact that hostile expression was classified in Arabic but not classified as hostile expression in Hebrew." In other words - classifying hostile expressions in Arabic as being identified with certain terrorist organizations .

Mark Zuckerberg, CEO of Meta, photo: AP

Another claim raised in the report is that Meta did not employ enough content reviewers to address the sharp increase in content problems (inciting posts, for example) published in it. In addition, it was claimed that Meta removed the possibility of using the hashtag "Al-Aqsa" (because it was classified as "al -Aqsa"), Meta users reported a reduction in content circulation, anti-Semitic content that violated Meta policy was not removed, and journalists' WhatsApp accounts were inadvertently disabled as a result of enforcement activity against terrorist groups.

The al-Aqsa hashtag has been removed, photo: Reuters

In response to the allegations in the report, Meta shared that its policy covers (deals with) words of praise for violence concerning dangerous organizations and people, and that it intends to act in order to prevent potential actual harm concerning the content. for interpersonal violence and definitions of words of praise, support and representation of dangerous organizations".

The Gaza Strip at the Walls, photo: E.P

"We have begun a policy development process in order to examine our definitions of praise, support and representation in our Digital Object Identifier policy, as well as to explore possible related approaches to social and political discourse."

Regarding the marking and disqualification of accounts and posts, she said: "We are working on ways to make the user experience of our disqualifications simpler and more transparent. We provide reasons when content is removed, in most cases."

Attacks on Hamas cyber targets during Operation "Guardian of the Walls" // Photo: IDF spokesman

Regarding content in Arabic that is enforced by "excessive enforcement" that may be found to be in violation of the policy: "We are already examining the routing of Arabic content by dialect. We have conducted research in order to improve our approach to routing content in a wide variety of Arabic dialects."

On anti-Semitic, Islamophobic and anti-Arab content: "We are committed to monitoring the overall prevalence of hate speech on our platform, however we face challenges measuring the prevalence at this level of detail."

were we wrong

We will fix it!

If you found an error in the article, we would appreciate it if you shared it with us