> Human Rights Watch also found repeated inaccurate application of the “adult nudity and sexual activity” policy for content related to Palestine. In every one of the cases, we reviewed where this policy was invoked, the content included images of dead Palestinians over ruins in Gaza that were clothed, not naked. For example, multiple users reported their Instagram stories being removed under this policy when they posted the same image of a Palestinian father in Gaza who was killed while he was holding his clothed daughter, who was also killed.
... so no nudity, no sexual activity
> For example, a Facebook user post that said, “How can anyone justify supporting the killing of babies and innocent civilians…” was removed under Community Standards on “bullying and harassment.” Another user posted an image on Instagram of a dead child in a hospital in Gaza with the comment, “Israel bombs the Baptist Hospital in Gaza City killing over 500…” which was removed under Community Guidelines on “violence and incitement.”
... arguing _against_ violence is flagged as harassment. Image and statement about Israeli action flagged as incitement
> In one case, an Instagram user received a warning that the comment she posted “may be hurtful to others.” The comment, which Human Rights Watch reviewed, consisted of nothing more than a series of Palestinian flag emojis.[77] In other cases, Meta hid the Palestinian flag from comment sections or removed it on the basis that it “harasses, targets, or shames others.”[78]
...that's just kinda indefensible.
> Many users reported posts on Instagram being removed when they criticized the Israeli government, including the leadership of Prime Minister Benjamin Netanyahu, no matter how nuanced or careful their posts were. Meta removed these posts under its “Dangerous Organizations or Individuals” and hate speech rules, respectively.
... which is crazy since there have been plenty of reasons to criticize Netanyahu well before the Oct 7 attacks.
On the topic of "From the river to the sea", they specifically mention that posts with this phrase, as well as others
> such as “Free Palestine,” “Ceasefire Now,” and “Stop the Genocide,” were repeatedly removed by Instagram and Facebook under “spam” Community Guidelines or Standards without appearing to take into account the context of these comments.
I.e. Meta's own cited policy was not that it "is considered hateful" as you say, but that this political expression was labeled as spam. That sounds pretty bogus.
As for the "is considered hateful by many", that's true, but is also subjective, and very tangled given that prominent Israeli conservatives have used very similar phrasing. https://en.wikipedia.org/wiki/From_the_river_to_the_sea#Simi...
If someone actually praising the Oct 7 attacks or something gets flagged/banned/etc, then sure, Meta is just applying its policies. That does not seem to be what's happening here. It's true that the report doesn't give numbers for these specific treatments, and the authors have a limited and biased view based on who responded to their call for evidence. But the limited view doesn't seem to be of Meta even-handedly and competently applying their policies.
_However_, I don't want to attribute everything to malice where incompetence may be at least a partial explanation. If motivated 3rd parties are repeatedly flagging a photo of a clothed, dead Palestinian as "nudity", and Meta's systems are built with the assumption that such user-provided flags are presumed trustworthy, then we could easily see how someone posting that photo would get incorrectly/inappropriately flagged.