Online watchdog prods platforms to tackle posts denying sexual assault on Oct. 7
Antisemitism monitor CyberWell says social media moderators are not doing enough to remove content that casts doubt on credible accounts of rape and other Hamas atrocities
Sharon Wrobel is a tech reporter for The Times of Israel.
On the front page of the March 27, 2024, edition of the New York Times, alongside stories about the collapse of the Francis Scott Key Bridge in Baltimore, access to abortion pills and Donald Trump’s court cases, was the headline “Israeli Hostage Tells of Sexual Assault in Gaza.”
Over approximately 4,000 words, freed hostage Amit Soussana recalled her harrowing kidnapping from Kibbutz Kfar Aza on October 7 and being forced to perform sexual acts on the man guarding her in Gaza.
“He sat me on the edge of the bath. And I closed my legs. And I resisted. And he kept punching me and put his gun in my face,” Soussana was quoted telling The Times. “Then he dragged me to the bedroom.”
Her account, which the paper noted was consistent with what she had told professionals immediately upon her release after 55 days in captivity, confirmed what other freed hostages, massacre survivors, family members and forensic experts had long said: Sexual violence, including rape and mutilation, had taken place on October 7 as Hamas-led terrorists brutally rampaged through southern Israel, and those abducted into Gaza continued to be subjected to sexual assaults.
Yet nearly eight months after October 7, and over two months after Soussana became the first to come forward about the abuses she suffered, accounts of sexual violence continue to be doubted or worse. This is despite mounting evidence that Hamas has used sexual assault as a weapon of war, including a documentary by former Meta chief operating officer Sheryl Sandberg on the systematic sexual violence, and a UN report that found “reasonable grounds” to back the allegations.
Thanks to lax enforcement or outdated standards, social media platforms have largely allowed Hamas apologists, Israel critics and others to spread and bolster the false narrative that accounts of sexual assaults, gang rape and other atrocities are either made up or wildly inflated, according to CyberWell, a nonprofit founded in May 2022 to create an open database to monitor and help remove antisemitic online content.
“Despite Hamas terrorists documenting their atrocities and livestreaming and uploading videos and photos onto social media platforms, extremists on social media quickly began denying the very fact of the sexual assault – narratives that gained traction and continue to be spread online today,” said CyberWell CEO Tal-Or Cohen Montemayor. “The denial of rape is an attempt to rewrite history – to obfuscate the deliberate crimes committed against women and redirect sympathy away from the victims and toward justification and celebration of their attackers.”
The unprecedented events of October 7 brought to the surface a wave of terror using the power of social media to aim at victims’ loved ones, harm Israelis and reach millions of viewers online in the Middle East and around the world.
David Saranga, the Foreign Ministry’s head of digital diplomacy, warns that social networks have become a strategic threat to the country and other Western-style democracies.
“No democratic country has the effective means to fight this culture of lies,” said Saranga. “Even if thousands report the tweet as harmful and it is eventually removed from the platform, the damage has already been done and millions have been exposed to the lie.”
A recent CyberWell analysis found that 135 social media posts in English and Arabic denying that Hamas perpetuated sexual violence and rape on October 7 had managed to reach more than 15 million users. Nearly half of the posts appeared on X, formerly Twitter, while 27% were on Facebook, 13% were on TikTok, and 6% were on Instagram.
CyberWell found that social media users denying sexual assault had taken place commonly pointed to the lack of hard evidence or testimony from rape survivors — many of whom were also murdered. As harrowing testimonies were made public, critics tried to poke holes in their credibility, accusing them of lying, and attempted to discount the reliability of journalists who reported on sexual assault.
In some cases, deniers have pointed to questionable claims regarding atrocities, made in the confusion and chaos immediately following the unprecedented massacre, to absolve Hamas of guilt.
Montemayor said some also maintain that Hamas members could not have committed sexual offenses as they are driven by religious Muslim ideology, alleging that victims were raped by Israelis, a conspiracy theory that dovetails with the mostly fringe idea that Israel was behind the October 7 atrocities.
Even though these narratives violate the policies of the social media platforms hosting this content, the average removal rate by the custodians or moderators of these platforms was only about 22%, lower than the 32% average removal rate for antisemitic posts in 2023, according to CyberWell.
Content moderators removed just over 24% of flagged posts on Facebook, 20% on YouTube, 12.5% on TikTok, and X labeled 4% of the tweets under the platform’s “limited visibility” flag and removed only 1.5% of posts.
Instagram, which is owned by Facebook parent company Meta, showed the lowest removal rate of October 7 rape denial content, with a zero percent action rate, according to the CyberWell report.
The low removal rate is due either to major gaps in enforcement of platform policies or to failure to include the October 7 massacre in the “list” of recognized violent events, the report said.
“Platforms must enforce their existing hate speech and sexual violence policies, recognize denial of October 7 sexual assault as prohibited content, and remove these posts at scale,” Montemayor demanded. “We are beyond seven months since October 7, and we are seeing that social media platforms are still not removing this content at scale and have not come out with any kind of statement or position on this content.”
Montemayor lamented that there is no automatic removal guidance for content moderators within the social media platforms, with the exception of pornography, in particular child pornography, and copyright violations as there are clear policies that such content is recognized as illegal activity.
“What they are doing is they are leaving it to a third-party fact checker system, meaning anything that is reported to social media goes to third-party fact-checking if it’s an issue of a matter of fact, and therefore it slows up the entire process it takes to verify or deny a claim while potentially false information remains online,” said Montemayor. “This results in an unprecedented amount of traction for the October 7 denial campaign.”
CyberWell aims to drive the enforcement and improvement of digital policies and community standards across the social media space and to combat online antisemitism and hate speech. It has used AI to build a real-time database, open to the public, that uses open-source intelligence to monitor and flag online antisemitism.
CyberWell is included in Meta’s trusted partner program, allowing it to communicate directly with Facebook and Instagram regarding content that it believes could be or is considered hate speech. It participates in a similar program with TikTok and also shares data with Elon Musk’s X upon finding relevant hashtags or spikes in antisemitism, Montemayor added.
The nonprofit has created a proprietary lexicon that was developed to flag highly likely antisemitic content online. The lexicon is rooted in the International Holocaust Remembrance Alliance, or what is known as the IHRA working definition of antisemitism. The flagged content is then manually vetted by CyberWell’s research team of four people and packaged into a high-level compliance alert for content moderators, policy teams and engineers of social media platforms.
“Effectively we are an online antisemitism compliance tool for social media platforms because we provide them with real data about how their own rules were failing to be enforced,” Montemayor remarked. “This empowers their teams to do the investigations independently, and then remove that content.”
Since the October 7 onslaught, CyberWell has helped to remove well over 50,000 pieces of content that violate social media platforms’ policies, Montemayor said.
However, social media platforms are still not removing such content at scale, she charged.
When contacted by The Times of Israel, TikTok acknowledged that some of the posts on its platform flagged in the CyberWell report were “prohibited.”
“This content is strictly prohibited and removed from TikTok, and we are grateful to our partners at CyberWell who help us continually strengthen how we protect our platform and combat antisemitism online,” a TikTok representative said in an e-mailed response.
Meanwhile, representatives from Meta, the parent company of Facebook and Instagram, did not respond to requests for comment before this article went to press. A representative for X could not be reached.
What needs to happen, according to Montemayor, is for major social media platforms to issue a statement declaring that October 7 denial is “prohibited content because it is a violent event denial meant to victimize victims of a major terrorist attack.”
CyberWell has called on all social media platforms to recognize and treat content denying and distorting the atrocities of October 7 the same way they would treat content denying or distorting the Holocaust, under policies restricting speech denying the existence of well-documented violent events.
“Social media platforms in the beginning were hesitant to make any kind of policy around the issue of October 7, as information continued to come out about the events,” Montemayor said. “But at this point, the hesitance to answer this call and to recognize that this is a purposeful antisemitic campaign and move it into an issue of policy is laziness that will result in violence against Jewish people and it’s unacceptable.”
