5 ms·
I say this not in defence of the Israelis (ab)use of Meta's takedown system, nor in defence of the Israelis conduct towards Palestinians: this is low-quality, s
by wk_end 1y ago
I say this not in defence of the Israelis (ab)use of Meta's takedown system, nor in defence of the Israelis conduct towards Palestinians: this is low-quality, sensationalist work.
ICW claims to be an "organization of independent journalists", but typos and grammatical errors are rampant, which I wouldn't expect from actual professional writers - unless the term "journalist" is being abused here.
Along those lines, serious journalists would not attempt to blackmail a multibillion dollar company like this:
> The motivation behind these leaks is the following: Stop all involvement with the Israeli government and their current genocide in Gaza. Until then, more leaks will be dropped. With each leak exposing a different aspect of corruption from censorship, to AI, to financial crimes.
The actual contents of the "leaks" aren't really new - the Israeli government has been making sweeping takedown requests to Meta to suppress material related to the war in Gaza. This was already well-known - see, for instance, HRW's report on this from 2023, itself cited in this report.
What makes this sensationalist is the conspiratorial thinking that follows and is embedded throughout.
There's certainly something interesting and dystopian about how Meta uses machine learning to extrapolate from successful human-verified takedown requests to begin automating acceptance of takedown requests, and how this creates, as an inevitable consequence, a kind of "censorship machine". But this is framed (without evidence) as a kind of "data poisoning" conspiracy - the Israeli intelligence agencies and perhaps Meta in tandem working to deliberately ensure this censorship is automated. When, of course, even if Meta wasn't using machine learning to automate this process, the Israeli agencies would almost certainly still be issuing these sweeping requests. And, conversely, with or without the Israeli takedown requests, there's no doubt Meta would be using ML to automate the takedown process at their scale. No conspiracy is necessary - the authors might do well to read Manufacturing Consent.
Even more absurdly, they "hypothesis [sic]" that "Israeli government [sic] must have insiders at Meta’s integrity organization in the form of individual contributor engineers who advised the Israeli cyberunit on how to abuse the content moderation system."; no evidence is presented for this claim.
Silly accusations and inferences like this are strewn throughout the paper, such as when it expresses shock that the Israeli government uses a form letter to submit the takedown requests.
It also veers outright into "10/7 Truther" territory:
> Throughout this reporting dataset, we see massive drops in reports on every 7th day. This of course corresponds to Shabbat or Saturday, the day which Jews refrain from work activities. However, October 7 2023 is also on Shabbat, so it's interesting to see that on this day of rest, and facing an overwhelming attack: 1. Both the IDF cyberunit division and Israeli attorney’s office were prepared and collaborating on that day to message Meta out of all organizations. 2. The IDF cyberunit division already had developed a new strategy of censoring countries that are not even involved in the attack like Iraq, Lebanon, Saudi Arabia, Egypt, Jordan.
The obvious explanation for this is that it's precisely because they were "facing an overwhelming attack" that the Israelis were working on the weekend. It's not surprising in the slightest that they had emergency procedures, either.
The HRW report had its own issues, but it was at least a serious attempt by serious people to understand the scale of the Israeli operation. By comparison, this is amateur work by amateurs.
- myrmidon 1y agoThis is harsh, but I agree with most of your points. The "blackmail" remark I find particularly damning: Threatening future leaks until Facebook stops "involvement with the Israeli government" is basically a self-admission that you put political success over journalistic integrity. That turns you from a "legitimate" whistleblower (like Snowden) into an activist (at best) in my eyes.
- ath92 1y agoJust out of curiosity, what do you believe Snowden intended by leaking what he leaked? From your comment it almost reads as if you think the publication was the point, not that those implicated by the leaks should be made to change their ways?
- myrmidon 1y agoSnowden himself says that he saw a surveillance state being built and the people responsible lying about it under oat in front of democratic representatives (see James Clapper). In his words, he wanted the American public to know about what was going on, and to get a say. But when you say "I'll stop publishing once my political goals are achieved", you are openly admitting that: 1) You publishing this informations causes harm 2) You are using that harm to achieve your political goals 3) You can live with the actual information/leaks staying hidden once the (possibly only tangentially related) goals are achieved (which indicates that the whole thing was primarily about achieving your political goals, not about getting out the information). That, to me, is the difference between "whistleblowing" and "blackmail".
- ath92 1y ago> 1) You publishing this informations causes harm …Harm to a company supporting genocide > 2) You are using that harm to achieve your political goals Snowden was also achieving political goals with his publishing > 3) You can live with the actual information/leaks staying hidden once the (possibly only tangentially related) goals are achieved (which indicates that the whole thing was primarily about achieving your political goals, not about getting out the information). Stopping once Meta stops the censorship would cause people to learn about what’s happening in Gaza, which I’m guessing is more important to the authors than exposing Meta’s censorship machine.