Elon Musk, chief executive officer of SpaceX and Tesla and owner of X/Twitter, gestures as he attends the Viva Technology conference dedicated to innovation and startups at the Porte de Versailles exhibition centre in Paris, France, June 16, 2023. Photo: REUTERS/Gonzalo Fuentes

It was half past midnight in Los Angeles on Saturday, October 7, already 10:30 am in Israel, when I was contacted by a senior representative of a major social media platform.

“I don’t know if you and your team have started looking at this content, but we need your help.”

My Israeli friends had already alerted me to Hamas’ horrific invasion, and for the next two hours, I frantically patrolled and reported the explosion of graphic content I saw on X — the platform with the least robust content moderation infrastructure.

I found a nauseating mass of tweets celebrating the massacre at the Nova music festival; videos of Hamas terrorists parading a young German woman, Shani Louk, naked and bleeding in the back of a truck while they spit on her; a young Israeli soldier being pulled from a car, bloodied but alive, then lynched by a crowd shouting “Yahud” — the Arabic word for Jew. For the record, X removed most of the content about 12 hours after I reported it.

My team and I from CyberWell — the world’s first open database dedicated to fighting antisemitism across social media — catalogue and study hateful content every day, but this was like nothing we’d ever seen.

It’s difficult to communicate what Israel and its people are suffering. Hamas’ onslaught has been compared to events like 9/11 and the Yom Kippur War, but the horror defies comparison for a very specific reason.

What makes this conflict unique is Hamas’s weaponization of social media and chat platforms for psychological warfare.

While some disturbing content is shared by terror victims and their families, desperately trying to communicate with loved ones — or find them amongst burnt and decapitated corpses — most of it is shared by terrorists and their allies for the express purpose of intimidating and threatening Jews around the globe.

As we know well, social media discourse directly affects the real world. Following Hamas’s call for a Day of Rage on Friday, October 13, many Jewish schools canceled classes, limited outdoor recess time, or told students not to wear uniforms that would identify them as Jews.

Parents were advised to keep their children away from social media in response to threats that Hamas would livestream footage of the 241 Israeli hostages to their personal social media accounts. In doing so, Hamas extended their terrorism beyond Israel and into the worldwide web.

Other extremist groups have surely noticed this tactic, and unless social media platforms act quickly, digital terror could become a worldwide security threat. Social media platforms were as surprised as the IDF by the October 7 attacks, and content moderation teams have been scrambling to remove and de-amplify the explosion of graphic content.

This is why robust monitoring and effective content moderation is so crucial. If effective steps had already been taken to curb antisemitism and violent Jew-hatred online, the amplification of terror during this crisis could have been mitigated from the start.

Given that, the unfolding hostage situation, and current realities on the ground, Trust & Safety teams are likely to be stretched thin for the foreseeable future. The data these teams and CyberWell collect and vet during this war will be vital to help expand their capacity for large-scale, rapid-response and automated content moderation.

This week we are releasing a compilation of every compliance alert we’ve sent to Facebook, Instagram, X, and TikTok since the beginning of the war.

Our AI-based monitoring detected a 50% increase in antisemitic content across those platforms since October 7. Incidence of the hashtag #hitlerwasright on X soared 29,000% in Arabic and 1,600% in English following the attack, and a Palestinian imam’s sermon from 2020 that exhorted the Arab world to “PUBG” the Jews — a call to genocide referencing an online game — was repackaged and distributed to the imam’s 20 million followers across all platforms.

Focusing in on how hate spreads is the key to effectively eliminating it, which is only possible through high-integrity data. Footage of the mass attack on Israeli civilians could, for example, be used as an initial dataset to train AI tools to more effectively identify this content and prevent it from being broadcasted to promote terror.

Regulators have an important role as well. Four days into the conflict, the European Union warned X and Meta for their failure to comply with the Digital Services Act (DSA), which allows the EU to fine companies up to 6% of global turnover if they do not systematically remove illegal hate speech and terrorist content.

Without similar legislation in the United States, however, platforms will have little incentive to make the necessary investments in user safety. Social media platforms and big tech should be regulated like other heavily supervised public industries, such as banking, pharmaceuticals, or waste management.

We should consider requiring a threshold of demonstrable automatic image and audio analysis to prevent the further glorification of violence — mandating that large social media companies and big tech companies invest a minimum amount of money for automated prevention of hate and violent content.

Ultimately, the role that social media platforms played in terrorizing the entire Jewish and Israeli population during and since the October 7 attack should give all of us pause. Antisemitic incidents were already at historic highs; they have increased further.

Platforms have been negligent in handling their moral responsibility to protect their users, from a lacking infrastructure to inexcusably slow response times. It’s time to impose legal responsibilities to close the gap and prevent our social media platforms from being hijacked by terror groups in the future.

Tal-Or Cohen is the Founder and Executive Director of CyberWell.