Dear John,

Just days ago, the ADL Center for Technology and Society launched something really big...

Our new Online Hate Index (OHI) is the first tool to detect and measure antisemitism on social media. It uses artificial intelligence that has been trained by experts in antisemitism and by trained volunteers.

OHI offers a clear picture of the state of hate on social media. We will use this tool to hold social media platforms accountable for how well they proactively take down hate and how well their content moderators respond to reports.

How big is the problem? How much antisemitism is there on social media?

In just a single week’s worth of Twitter and Reddit samples that we used OHI to analyze, we estimate the potential reach of antisemitic tweets in that week was 130 million people.

In this initial study, both Reddit and Twitter failed to remove over 70 percent of the antisemitic content ADL uncovered, even after the posts had been up for nearly a month. Then, after ADL reported this hateful content and more than two months after ADL first found it, 56 percent of antisemitic Reddit comments and 57 percent of antisemitic tweets remained on the platforms.

That’s a lot of eyes on antisemitic content on two of the most influential social media platforms. The danger is that hateful content can normalize violent extremism, enable extremist radicalization and recruitment, promote dangerous conspiracy theories, and, in some cases, lead to deadly violence.

We should make clear that this OHI analysis is not an indictment of Reddit and Twitter. In fact, we could only do this analysis because of these companies’ commitments to transparency and data-sharing with third parties. Some platforms, such as Facebook, have even gone so far as to hide what they do know about who is targeted by hate on their platform from their own auditing teams and other efforts at accountability. While Reddit and Twitter have far more to do, they have made substantial recent strides.

How Does OHI Work?

For the first time, we are using the combined powers of artificial intelligence and ADL’s expertise to uncover antisemitic content online at scale, able to view thousands of pieces of content a second. OHI relies on our antisemitism classifier, an algorithm that learns to recognize antisemitism by being given numerous examples of both offensive and innocuous content.

This all builds on the input of experts on antisemitism and a trained group of volunteer labelers who identify as Jewish. They read through thousands of social media posts to identify antisemitic content. We acknowledge that viewing antisemitic content can have harmful effects on our volunteer labelers, so we encourage frequent breaks and provide support from ADL staff if things become overwhelming.

Taking Action

If you’d like to join the volunteers involved in this important project to fight online antisemitism, and you have a familiarity with social media, please click here to take a short demographic survey. This will help our team to connect with the volunteers we need for this round of the project, as well as to connect you to future OHI work that detects other identity-based forms of hate.

To better address antisemitic content online, ADL developed recommendations for both platforms like Twitter and Reddit and for the government, the latter as part of our sweeping REPAIR plan to decrease hate online through regulation and reforms.

OHI is an innovative new tool to spot, report and ultimately disrupt online antisemitism and other forms of hate and to hold social media platforms accountable for the content they host on their influential platforms. We are very appreciative of the volunteers for their role in this project, and to everyone who speaks out against online hate with ADL.

Sincerely,

Jonathan Greenblatt

CEO and National Director

ADL