"Israeli Group Claims It’s Working With Big Tech Insiders to Censor “Inflammatory” Wartime Content", 10 January 2024

A small group of volunteers from Israel’s tech sector is working tirelessly to remove content it says doesn’t belong on platforms like Facebook and TikTok, tapping personal connections at those and other Big Tech companies to have posts deleted outside official channels...

...Iron Truth claims its tech industry back channels have led to the removal of roughly 1,000 posts tagged by its members as false, antisemitic, or “pro-terrorist” across platforms such as X, YouTube, and TikTok...

The project’s public face is a Telegram bot that crowdsources reports of “inflammatory” posts, which Iron Truth’s organizers then forward to sympathetic insiders...

So far, nearly 2,000 participants have flagged a wide variety of posts for removal, from content that’s clearly racist or false to posts that are merely critical of Israel or sympathetic to Palestinians, according to chat logs reviewed by The Intercept...

While Iron Truth appears to have frequently conflated criticism or even mere discussion of Israeli state violence with misinformation or antisemitism, Kaganovitch says his views on this are evolving. “In the beginning of the war, it was anger, most of the reporting was anger,” he told The Intercept. “Anti-Israel, anti-Zionist, anything related to this was received as fake, even if it was not.”...

Iron Truth’s founder declined to share the names of its “insiders,”...In general, “they are not from the policy team but they have connections to the policy team,” Kaganovitch told The Intercept, referring to the personnel at social media firms who set rules for permissible speech. “Most of them are product managers, software developers. … They work with the policy teams with an internal set of tools to forward links and explanations about why they need to be removed.” While companies like Meta routinely engage with various civil society groups and NGOs to discuss and remove content, these discussions are typically run through their official content policy teams, not rank-and-file employees...

Kaganovitch said the project has allies outside Israel’s Silicon Valley annexes as well. Iron Truth’s organizers met with the director of a controversial Israeli government cyber unit, he said, and its core team of more than 50 volunteers and 10 programmers includes a former member of the Israeli Parliament...

“Across our products, we have policies in place to mitigate abuse, prevent harmful content and help keep users safe. We enforce them consistently and without bias,” Google spokesperson Christa Muldoon told The Intercept. “If a user or employee believes they’ve found content that violates these policies, we encourage them to report it through the dedicated online channels.” Muldoon added that Google “encourages employees to use their time and skills to volunteer for causes they care about.” In interviews with The Intercept, Kaganovitch emphasized that he works on Iron Truth only in his free time, and said the project is entirely distinct from his day job at Google.

Meta spokesperson Ryan Daniels pushed back on the notion that Iron Truth was able to get content taken down outside the platform’s official processes, but declined to comment on Iron Truth’s underlying claim of a back channel to company employees. “Multiple pieces of content this group claims to have gotten removed from Facebook and Instagram are still live and visible today because they don’t violate our policies,” Daniels told The Intercept in an emailed statement. “The idea that we remove content based on someone’s personal beliefs, religion, or ethnicity is simply inaccurate.” Daniels added, “We receive feedback about potentially violating content from a variety of people, including employees, and we encourage anyone who sees this type of content to report it so we can investigate and take action according to our policies,” noting that Meta employees have access to internal content reporting tools, but that this system can only be used to remove posts that violate the company’s public Community Standards.

Neither TikTok nor LinkedIn responded to questions about Iron Truth. X could not be reached for comment...

Iron Truth has routinely used the fog of war as evidence of anti-Israeli disinformation.

At the start of the project in the week after Hamas’s attack, for example, Iron Truth volunteers were encouraged to find and report posts expressing skepticism about claims of the mass decapitation of babies in an Israeli kibbutz...

Similarly, Iron Truth’s early efforts to take down “disinformation” about Israel bombing hospitals now contrast with weeks of well-documented airstrikes against multiple hospitals and the deaths of hundreds of doctors from Israeli bombs...

While most of the links no longer work, several are still active. One is a video of grievously wounded Palestinians in a hospital, including young children, with a caption accusing Israel of crimes against humanity...

While it’s possible for Meta posts to be deleted by moderators and later reinstated, Daniels, the spokesperson, disputed Iron Truth’s claim, saying links from the list that remain active had never been taken down in the first place. Daniels added that other links on the list had indeed been removed because they violated Meta policy but declined to comment on specific posts...

“It’s really hard to identify disinformation,” Kaganovitch acknowledged in an interview, conceding that what’s considered a conspiracy today might be corroborated tomorrow...

Throughout October, Iron Truth provided a list of suggested keywords for volunteers in the project’s Telegram group to use when searching for content to report to the bot... Some of those phrases — including “Israel bomb hospital”; “Israel bomb churches”; “Israel bomb humanitarian”; and “Israel committing genocide” — were suggested as disinformation keywords as the Israeli military was being credibly accused of doing those very things...

Even the words “Israel lied” were suggested to Iron Truth volunteers on the grounds that they could be used in “false posts.” On October 16, two days after an Israeli airstrike killed 70 Palestinians evacuating from northern Gaza, one Telegram group member shared a TikTok containing imagery of one of the bombed convoys. “This post must be taken down, he is a really annoying liar and the amount of exposure he has is crazy,” the member added. A minute later, the Iron Truth administrator account encouraged this member to report the post to the Iron Truth bot...

Telegram transcripts show the group’s administrator frequently encouraged users to flag posts accusing Israel of genocide or other war crimes...

Iron Truth has found targets outside social media too. On November 27, one volunteer shared a link to NoThanks, an Android app that helps users boycott companies related to Israel. The Iron Truth administrator account quickly noted that the complaint had been forwarded to Google. Days later, Google pulled NoThanks from its app store, though it was later reinstated.

The group has also gone after efforts to fundraise for Gaza. “These cuties are raising money,” said one volunteer, sharing a link to the Instagram account of Medical Aid for Palestinians. Again, the Iron Truth admin quickly followed up, saying the post had been “transferred” accordingly...

Several weeks after its launch, Iron Truth shared an infographic in its Telegram channel asking its followers not to pass along posts that were simply anti-Zionist. But OCT7, an Israeli group that “monitors the social web in real-time … and guides digital warriors,” lists Iron Truth as one of its partner organizations, alongside the Israeli Ministry for Diaspora Affairs, and cites “anti-Zionist bias” as part of the “challenge” it’s “battling against.”

Despite Iron Truth’s occasional attempts to rein in its volunteers and focus them on finding posts that might actually violate platform rules, getting everyone on board has proven difficult. Chat transcripts show many Iron Truth volunteers conflating Palestinian advocacy with material support for Hamas or characterizing news coverage as “misinformation” or “disinformation,” perennially vague terms whose meaning is further diluted in times of war and crisis...

Social media accounts of Palestinian journalist Yara Eid; Palestinian photojournalist Motaz Azaiza; and many others involved in Palestinian human rights advocacy were similarly flagged by Iron Truth volunteers for allegedly spreading “false information.”

Iron Truth has at times struggled with its own followers. When one proposed reporting a link about Israeli airstrikes at the Rafah border crossing between Gaza and Egypt, the administrator account pointed out that the IDF had indeed conducted the attacks, urging the group: “Let’s focus on disinformation, we are not fighting media organizations.” On another occasion, the administrator discouraged a user from reporting a page belonging to a news organization: “What’s the problem with that?” the administrator asked, noting that the outlet was “not pro-Israel, but is there fake news?”

But Iron Truth’s standards often seem muddled or contradictory. When one volunteer suggested going after B’Tselem...the administrator account replied: “With all due respect, B’Tselem does publish pro-Palestinian content and this was also reported to us and passed on to the appropriate person. But B’Tselem is not Hamas bots or terrorist supporters, we have tens of thousands of posts to deal with.”...

The group’s founder says that Iron Truth leadership have met with Haim Wismonsky, director of the controversial Cyber Unit of the Israeli State Attorney’s Office...

Wismonsky did not respond to a request for comment...