[icw_nru]

Human rights watch's report covers more individual examples.

To quote: "Of the 1,050 cases reviewed for this report, 1,049 involved peaceful content in support of Palestine that was censored or otherwise unduly suppressed, while one case involved removal of content in support of Israel."

This leak aims at looking at the bigger picture across all of Meta's 3 billion users.

Of course, Meta can chose examples of actually violating posts removed and show that as counter proof, or even posts that are violating that are not yet removed. But anyone familiar with how ML models work knows that false positives / false negatives exists.

Its the degree to which the ML models primarily censor almost any content related to Israel/Palestine, the systemic nature of targeting specific countries, such as Palestine, Egypt, Jordan, and the fact that per-capita, Israel is the country that most abuses the content enforcement system (3x more than any other country).

[notavalleyman]

> Of course, Meta can chose examples of actually violating posts removed and show that are counter proof, or even posts that are violating that are not yet removed

No, meta don't need to prove anything to anyone.

It's you who alleges that the content should have stayed up, so what's your evidence?

You're telling me I need to go and read a HRW pdf instead? Okay where is that?

[icw_nru]

The report: https://www.hrw.org/report/2023/12/21/metas-broken-promises/...

Unfortunately links to individual posts can't be accessed as the posts themselves are removed. The HRW report is excellent as they documented this individual cases and recorded them.

[notavalleyman]

Okay well you don't have any proof, any in general, I would consider it a good use of tax shekels to reduce the number of pro-hamas posts on social media. So until you can dig up any proof, I'm considering this whole post to be a nothing burger

[icw_nru]

Did you look at the article? This investigation directly corroborates existing reports from third parties like Human Rights Watch. There is even an intake form directly from the Israeli government calling for the censorship posts of posts at Meta. We even posted their phone and fax number in case anyone is interested in having a friendly chat with them.

All data collected is directly from Meta, and the whisteblowers themselves are open to sharing this data with any authority or court willing to look into this. Everything is well documented. Where and how the data was obtained is also documented as well.

Or alternatively, you can wait for the next leak.

[text0404]

google search for "human rights watch israel censorship" turned this up: https://www.hrw.org/sites/default/files/media_2023/12/ip_met...

check "illustrative examples" section.

[alangibson]

Wow, is almost like you're committed to obfuscation what Israel is doing. Couldn't possibly be that tho...

[strulovich]

All 1049 posts were peaceful? The pdf mentions this was mostly after October 7th, a terrorist (as in, meant to induce fear by targeting civilians) attack which was live streamed on Facebook and posted repeatedly during that day.

I’m surprised the Israelis are so capable with intelligence, yet bungled this so much that not one post they pointed out was violent?

I’m happy to stand corrected, but when someone shows a perfect record in a data review I’m naturally suspicious.

EDIT: I’m confusing the linked PDF and HRW’s report. But I still have doubts about HRW’s numbers.

[whatshisface]

In this case it's being alleged that sending a thousand false takedown requests which were then acted on would not be a bungling, but rather a success.