A 3D-printed miniature model of Elon Musk and the X logo are seen in this illustration taken Jan. 23, 2025. Photo: REUTERS/Dado Ruvic/Illustration

Recently, as Jewish Insider reported, bipartisan lawmakers in Congress hailed what they saw as a major advance in fighting online antisemitism — X’s new location feature.

The new tool, showing which country an account operates from — had started revealing that some accounts spreading antisemitic content in US political discussions were based overseas. For legislators on both sides, this represented a digital unmasking.

Rep. Don Bacon (R-NE) remarked that the feature exposed “foreign interests trying to spread antisemitic poison” while “masquerading as Americans.”

Sen. James Lankford (R-OK) insisted Americans “deserve to know which accounts are run from abroad, so we know the true source of these narratives.” Rep. Tom Suozzi (D-NY) took a geopolitical view: “Beijing, Moscow and Tehran know they cannot defeat us economically or militarily, so they exploit controversial issues, like Israel and antisemitism, and try to divide.” Former UN Ambassador Nikki Haley declared the feature “a huge win for transparency and American security.”

The story is appealing: foreign enemies weaponize antisemitism to fracture American unity, and transparency about account origins helps us counter these external threats. There’s truth here — bad actors do exploit divisive topics.

But this celebration reveals a dangerous misdiagnosis.

As the lead of the Decoding Antisemitism project — which has examined over 300,000 items of digital content across multiple crises — I’ve identified three distinct but connected drivers of online antisemitic radicalization: coordinated malicious actors (foreign and domestic), algorithmic amplification through platform design, and homegrown participatory dynamics enabled by online communication itself — anonymity, mutual reinforcement, and the normalization of extremism through constant exposure.

The issue isn’t that foreign influence exists — it does. The problem is treating it as the primary driver while overlooking the structural and domestic conditions that allow antisemitic narratives to take root and spread.

How we diagnose problems determines how we design solutions — and misdiagnosis doesn’t just limit our response, it actively redirects resources, attention, and political will away from factors we can actually control.

Comprehending online antisemitism demands a virological approach: examining not just where accounts originate, but how hate narratives evolve, which platform features enable transmission, and what conditions allow them to thrive. Yet social media platforms remain essentially black boxes — we lack systematic tools to understand dynamics unfolding within these digital spaces.

Lawmakers celebrate a feature revealing account geography while leaving the actual black box — algorithmic recommendations, engagement optimization, and content amplification — completely unexamined.

The Conflation Problem

The Jewish Insider article and quoted lawmakers collapse “foreign,” “adversarial,” and “antisemitic” into one category. This conflation obscures more than it reveals.

It treats geographic origin as definitive of intent and impact. An account in South Asia or Eastern Europe engaging with US politics isn’t necessarily a state-directed operation. It may simply be someone with opinions about American affairs. Account location reveals nothing about whether content is coordinated, conspiratorial, state-driven, or simply individual opinion.

More crucially, emphasizing foreign accounts distracts from what we know empirically about domestic antisemitic content production.

Following the October 7 attacks, antisemitic discourse surged to 36-38% of comments on major UK news outlet YouTube channels — nearly double the pre-crisis baseline. After the Washington museum shooting in May 2025, antisemitic content averaged 43% across major English-language news channels, with some reaching 66%.

These aren’t fringe platforms infiltrated by foreign bots — they’re mainstream digital spaces where domestic audiences actively produce and amplify antisemitic narratives.

Research on antisemitic discourse spread reveals a three-phase domestic process: elite figures make strategically ambiguous statements, digital intermediaries (podcasters, YouTubers, influencers) reframe and sharpen this messaging, and comment sections collapse ambiguity into explicit hate speech.

This “cascading radicalization” is primarily homegrown, driven by domestic actors and platform dynamics — not solely foreign infiltration.

Our analysis cannot definitively establish every anonymous commenter’s geographic origin. What we observe are linguistic and cultural markers — idiom, references, political framings — indicating domestic participation, combined with the absence of coordination patterns typical of bot networks. The antisemitic discourse we documented emerges through “dialogical warfare”: organic exchanges between users presenting as ordinary Americans who deploy antisemitism as an explanatory framework for complex issues.

When a US Congressmember amplifies antisemitic tropes, when popular podcasts platform guests trafficking in conspiracy theories about Jewish power, when partisan media deploy dual loyalty accusations — these aren’t foreign operations. They’re homegrown productions embedded in American political discourse and amplified through domestic networks.

The Missing Architecture

Most striking about celebrating X’s location feature is what remains unexamined: the platform itself.

There’s virtually no discussion about platform design, algorithmic amplification, recommendation systems, the attention economy, or structural dynamics allowing hateful content to scale. The feature is treated as inherently truth-revealing, exposing hidden foreign manipulation.

But this framing evades more important questions: why do certain narratives spread, how do platform architectures enable amplification, and how do online communication conditions — anonymity, mutual reinforcement, constant exposure to extremity — create environments where antisemitic ideas mutate and take hold?

The answer has little to do with account location and everything to do with how platforms are built. Engagement-based algorithms reward emotionally provocative content — outrage, fear, tribal solidarity. Recommendation systems create filter bubbles and radicalization pathways. Virality architecture privileges simplification, moral clarity, and villain identification. The attention economy systematically rewards polarizing, enraging content. These are design choices, not inevitable features.

Meanwhile, online communication conditions themselves — anonymity removing social accountability, mutual confirmation among like-minded voices, omnipresent hate speech normalizing extremity — create participatory environments where ordinary users become active radicalization contributors.

Contemporary antisemitism increasingly operates through coded expressions, memes, and multimodal signals evading simple keyword detection. The watermelon emoji, the paraglider symbol — these function as in-group markers regardless of geographic location. Strategic ambiguity, not foreign origin, enables antisemitic narratives to spread while maintaining plausible deniability.

The Political Convenience

The “foreign adversaries spreading antisemitism” narrative aligns with a bipartisan preference: attributing social breakdown to hostile external actors. This framing is politically convenient across the spectrum.

For Republicans, it allows condemning antisemitism without confronting how right-wing media has mainstreamed antisemitic conspiracy theories — “great replacement” narratives, George Soros accusations, “globalist” rhetoric. For Democrats, it enables criticizing online hate without reckoning with how segments of progressive activism have normalized anti-Zionist rhetoric often sliding into antisemitic tropes about Jewish power and loyalty.

The foreign influence frame permits symbolic accountability –the appearance of action without institutional change. Lawmakers can call for location transparency, celebrate platform implementations, and position themselves as defenders against external threats. What they needn’t do is examine how American political rhetoric contributes to normalizing antisemitism, push for regulatory interventions altering platform incentives, or confront how online communication creates radicalization pathways.

This isn’t analytical sloppiness. It’s moral abdication.

What Accountability Would Actually Require

Genuine accountability for online antisemitism requires confronting all three drivers — not just one.

First, acknowledging that while malicious actors (foreign and domestic) exploit divisive issues, they operate within a larger ecosystem. We must recognize the United States as an active site where antisemitic ideas are produced, circulated, and normalized through domestic political culture, media ecosystems, and participatory online dynamics — not merely as an innocent target.

Second, confronting how platform architecture shapes what spreads, and demanding transparency not just about account locations but about algorithmic recommendations, content moderation, the attention economy’s incentives, and metrics driving platform design.

Third, recognizing that high-profile domestic actors — politicians, media figures, influencers with millions of followers — bear far more responsibility for mainstreaming antisemitic narratives than anonymous accounts. We must understand the three-phase process through which elite ambiguity cascades into radicalized discourse.

Fourth, examining how partisan political discourse traffics in antisemitic tropes through strategic ambiguity — and being willing to call this out when politically inconvenient.

Fifth, acknowledging online communication conditions themselves — anonymity, mutual reinforcement, constant exposure to extremity — are creating environments where ordinary users become radicalization participants.

X’s location feature may provide useful information about one factor among many. But treating this as revelatory, exposing the “true source” of antisemitic narratives, is a dangerous misdiagnosis.

If democracies want to confront antisemitism seriously, they must address all three drivers: monitor and counter malicious actors where they exist, examine platform architectures amplifying hate, and confront participatory dynamics and communication conditions enabling antisemitic narratives to flourish in mainstream spaces.

How we diagnose problems shapes how we conceptualize solutions. Focusing exclusively on foreign actors may be politically convenient. It will not protect the public sphere.

Dr. Matthias J. Becker is a researcher in discourse studies and is affiliated with the Woolf Institute in Cambridge. He leads the Decoding Antisemitism project, which analyzes how antisemitic discourse spreads in digital communication.