Press Release | Supreme Court: Social Media Platforms Can Moderate Dangerous Content But Choose Not To

(February 28, 2023)

In oral arguments for Supreme Court cases Gonzalez v. Google and Twitter vs. Taamneh, social media giants argued that controlling what their algorithms recommend would require wholesale suppression of even unobjectionable content – essentially, that it would be too difficult to remove truly objectionable content without casting too wide a net.

CyberWell, the world’s first live database of online antisemitism, shows how inaccurate that is. With AI, human verification, and open-source intelligence technology, CyberWell collects and vets antisemitic content across social media, classifying it according to the community standards it violates and using the internationally accepted IHRA definition to identify precise antisemitic narratives.

Trained moderators with appropriate tools can enforce existing guidelines with precision – it’s just a matter of priority.

In May 2022, CyberWell found that platforms removed just 21% of verified antisemitic content. By year’s end, after sending alerts detailing the extent of the issue and demonstrating how such content could be identified and removed, the removal rate stood at less than 24%.

“When there is a legal or monetary penalty involved, social media companies have allocated appropriate resources to content moderation,” said CyberWell founder and extremism expert Tal-Or Cohen Montemayor, Adv. “Responding to companies who want to protect their intellectual property, advertisers who want to protect clean brands, and regulators who want to shut down illegal activity, they have been able to deploy extensive automated and human resources to control what their sites recommend. It’s time for them to do the same for antisemitism and other forms of hate.”

Twitter recently intervened in its own algorithm to boost the reach of Elon Musk’s tweets. They can intervene to limit the scope of harmful content, but are choosing not to.