Digital antisemitism and other forms of online hate often surge during geopolitical crises and waves of real-world violence.
The natural philanthropic impulse is to fund awareness.
We respond with billboard campaigns, educational curricula, and public solidarity statements. These are noble and vital efforts. But they are inherently reactive. If you are funding awareness while ignoring the physical plumbing of the internet, you are treating a burst pipe by wiping down the counters.
The modern reality of digital hate is that it is highly networked, technologically sophisticated, and aggressively operationalized. Fighting it requires an equal and opposite infrastructure.
At Fighting Online Antisemitism (FOA), we track how online vitriol mutates into physical danger every day. Understanding this fight requires shifting from an “awareness” mindset to an “operational” one. Digital hate is not an abstract cultural mood; it is a scalable software and systems problem.
1. The Operational Reality: Why Tech Submissions Fail
Many of us underestimate what it actually takes to scrub hate from the internet. There is a pervasive assumption that reporting a piece of content is a simple, automated action: you click a flag icon, an algorithm reviews it, and the content disappears.
It almost never works that way.
To achieve a high success rate — such as FOA’s benchmark of successfully removing 67% of monitored antisemitic content across major social platforms, as detailed in our 2025 Annual Report — requires an intense, human-led operational model. Nonprofits entering this space often fail because they underestimate three operational realities:
- The Compliance Chasm: Social media platforms do not respond to moral outcries; they respond to their own Terms of Service (ToS) and local legislation (like the European Union’s Digital Services Act). Removing a piece of content requires a trained analyst who can map a specific piece of hate speech directly to a platform’s specific legal and regulatory frameworks.
- The Linguistic Cat-and-Mouse Game: Hate actors do not use static slurs. They use “algo-speak” — coded language, deliberate typos, and visual memes designed to bypass automated content moderation filters. Popular AI moderation tools often assign very low toxicity scores to deeply antisemitic texts due to these sophisticated blind spots. Human eyes remain mandatory.
- The Scale Paradox: The volume of content is staggering. In our joint reporting with global monitoring partners, we highlighted a 400% surge in calls for terrorism and violence targeting Jews online following major crises, with an alarming shift where 18% of all monitored antisemitic material directly advocated physical harm. FOA relies on a global network of multilingual volunteers, backed by proprietary tracking software, to monitor platforms in real time. Without software to aggregate, triage, and categorize these threats, human teams are easily overwhelmed.
When philanthropy underfunds back-end infrastructure — the database architecture, volunteer coordination platforms, and analyst training — they leave nonprofits trying to defend a digital fortress with duct tape and good intentions.
2. From Pixels to Pickets: The Nova Festival Exploitation
The boundary between online chatter and physical threat no longer exists. For foundation program staff assessing risk, digital monitoring must be viewed as an early-warning system for physical violence.
A stark proof of concept occurred when our specialized monitoring unit identified a deeply concerning trend line inside a domestic white nationalist accelerationist cell operating on X (formerly Twitter). The group had begun using the moniker “Nova Now” — explicitly using the trauma of the October 7th music festival massacre as a tactical blueprint for violence on American soil. Our teams tracked the group’s evolution through a complex “rabbit hole” of nested threads. Over several months, we watched the language shift from abstract, generalized hate speech to highly operational specifics. Cell members began actively discussing the acquisition of close-quarters tactical weapons (knives) and explicitly stated it was “time for violence,” zeroing in on a countdown to April 1st — the first night of Passover.
Because our team operates with a strict data-collection infrastructure, we were able to secure a comprehensive evidentiary file, including the digital fingerprints and specific handles of the ringleaders. We transmitted this intelligence directly to the FBI Detroit Field Office’s hate crimes division via a confidential channel.
The result? The domestic terror conspiracy was completely disrupted months before its planned execution date. This intervention followed two previous disruptions of holiday-centric domestic terror plots in the same Michigan region. This reinforces a terrifying reality: the distance between an online post and a physical attack has effectively shrunk to zero.
The lesson for philanthropy is stark: Digital monitoring is a form of physical security. If a foundation funds the physical security of a community center by paying for guards and cameras, but ignores the digital monitoring that detects threats before they arrive at the front door, the funding strategy is fundamentally incomplete. Real-time digital tracking is the only way to get ahead of the news cycle.
3. The Donor Case: Funding Infrastructure, Not Just Initiatives
Why is this work chronically underfunded relative to its impact? The simple answer is that infrastructure is rarely glamorous. It is much easier to fund a high-profile summit or a viral video campaign than to fund the server architecture, data pipelines, and legal analyst hours required to keep hate off network servers.
We need a shift in donor psychology. Combating online antisemitism must be viewed through the same lens as funding civic infrastructure, public health data tracking, or legal defense funds.
The Recommendation for Funders: Shift from project-based, short-term grants to multi-year operational funding specifically earmarked for technology optimization, automated scraping tools, and platform compliance training.
When you fund a nonprofit’s ability to build sophisticated monitoring tools or train hundreds of volunteer monitors fluent in shifting digital subcultures, you are buying long-term capability. You are ensuring that when the next geopolitical crisis triggers a spike in online hate, the digital fire department already has the trucks, hoses, and water pressure ready to respond.
The Path Forward
The digital ecosystem is changing rapidly, driven by decentralized platforms and generative AI tools that make the mass production of hate content cheaper and faster than ever before.
At Fighting Online Antisemitism, our focus remains on the unglamorous, relentless work of digital sanitation — monitoring, reporting, and clearing the digital pathways so that hate cannot take root and blossom into physical violence. But nonprofits cannot hold the line using duct tape and good intentions. We need a philanthropic community willing to invest in the digital steel and concrete required to protect our shared online world.
