The two key takeaways—which confirm what Palestinian users and civil society—have been documenting and experiencing for years: 1) Meta's censorship of Palestinian content during #SaveSheikhJarrah and Israel's assault on Gaza in May 2021 had adversely violated people's rights.
2) The report confirms that Meta's content moderation policies and practice are biased against Palestinians and Arabic speaking users. For example, Meta had greater over-enforcement of Arabic content compared to Hebrew, which was largely unmoderated.
Another example BSR found was that Meta has classifiers to automatically detect hostile Arabic speech, but none for Hebrew. This is astounding given the escalating incitement to violence and calls to genocide directed at Palestinians online by Israelis. nytimes.com/2021/05/19/tec…
Mob Violence Against Palestinians in Israel Is Fueled by Groups on WhatsApp (Published 2021)
The report also found that Palestinians have had higher erroneous censorship rate under Meta's dangerous individuals and orgs (DOI) policy than their Israeli counterparts. One reason being, Meta's DOI policy disproportionality lists and targets Muslim individuals and entities.
BSR notes, though, these are instances of "unintentional bias"— a conclusion I strongly disagree with. This discriminatory system and censorship of Palestinian voices has existed for years. We have repeatedly raised with @Meta, with no action. The negligence is intentional.
The DOI policy in itself is embedded in systems of institutional racism and discrimination. Muslims are disproportionately designated as terrorists. Israelis? Meta doesn't even have classifiers for hostile speech in Hebrew. What's the company rationale behind this decision?
The report validates user experience of shadow banning in May 2021 (while not using this terminology). For example, users accumulated “false” strikes that impacted visibility and engagement after posts were erroneously removed for violating content policies.
In other words, users were put in Facebook and Instagram jails without their knowledge and even when their content was deleted in error. Since last year, Meta denied it had placed this type of penalty or restriction on Palestinian or pro-Palestinian users.
While the report validates what Palestinians and civil society have been stating for years. It's disappointing that it did not share any quantitative data on: 1) Content removal rates for Arabic vs. Hebrew content 2) Number of censorship requests sent by the Israeli cyber unit
These numbers are extremely important. Meta's lack of transparency in this regard is appalling at this point. We need transparency on how companies operate in contexts of conflict and occupation where existing power asymmetries have real life consequences on oppressed people.
Palestinian feminist made in Saudi | MENA Policy Manager @AccessNow | Policy Analyst @AlShabaka | ADHDer | Personal account & views
Missing some tweets in this thread? Or failed to load images or videos? You can try to .