The surge of anti-Semitic content across social media platforms since October 7, 2023, has occurred alongside an equally consequential collapse of Israel’s capacity to shape its own narrative online. Yet, the existing literature treats these as separate problems belonging to separate fields. This paper argues that they are the same problem: anti-Semitic content and Israeli public diplomacy content are ranked by the same engagement-driven algorithms, which reward emotional intensity over documentary accuracy regardless of which side supplies it. Drawing on platform monitoring data from the Institute for Strategic Dialogue (ISD), the Center for Countering Digital Hate (CCDH), and CyberWell; recent scholarship on algorithmic anti-Semitism; the long-running scholarly dispute over the “new anti-Semitism” and the International Holocaust Remembrance Alliance (IHRA) working definition; and reporting on the post-October 7 breakdown of Israel’s hasbara (state public diplomacy) apparatus, the paper traces how contested definitions, platform incentives, and state communication failure combine to produce a contest over legitimacy rather than a contest over fact. It concludes that regulatory instruments such as the European Union’s Digital Services Act govern platform conduct without resolving the prior definitional dispute over what counts as anti-Semitic, and that this unresolved dispute does more analytic work in the conflict than most policy literature acknowledges.
1. Introduction
Two claims circulate about social media and the Israeli-Palestinian conflict, and both are true; this is the analytic problem this paper addresses. The first is that anti-Semitic content online has surged since October 7, 2023, a trend documented so consistently across independent monitors that it leaves little room for dispute. The Institute for Strategic Dialogue (ISD), a London-based think tank that tracks online extremism, recorded a more than fifty-fold increase in the absolute volume of anti-Semitic comments on YouTube videos about the Israel-Palestine conflict in the three days after the Hamas attack, compared with the three days before it, and a 2.4-fold rise in the proportion of anti-Semitic comments relative to overall commentary (Rose et al., 2023). The Center for Countering Digital Hate (CCDH), a nonprofit based in the United States and the United Kingdom, reported that X was still hosting 85 percent of the anti-Semitic posts it had flagged to the platform in November 2023, a week after reporting them, and that Holocaust denial videos remained easy to find on TikTok despite the platform’s stated policies against them (Lavelle, 2023; Navarro, 2024). The American Jewish Committee’s survey work identifies online spaces as the setting where American Jews most often encounter anti-semitism. In its 2023 survey, 62 percent of respondents reported seeing or hearing anti-semitism online or on social media in the previous twelve months (American Jewish Committee, 2024).
The second claim is that Israel has lost, by the admission of its own commentators, the contest for global public opinion that Israeli officials call hasbara, a Hebrew word meaning “explanation” that has become the standard term for Israel’s state public diplomacy. This is not a claim manufactured by critics of Israeli policy. It appears in commentary by Israeli analysts published in the Jerusalem Post and in analysis from the American Foreign Policy Council, a Washington think tank, which describe hasbara as outdated and poorly suited to an information environment in which narratives built to appeal to emotion outperform narratives built to inform (Berman, 2024; Kahn, 2026; Siman-Tov, 2025). Israel’s Ministry of Public Diplomacy, re-established only at the start of 2023, was shut down by government decision on October 22, 2023, barely two weeks into the war; it might have been expected to help fight in the information domain, after its minister resigned, calling the ministry a waste of public funds (Jerusalem Post, 2023). That institutional decision itself became evidence of the coordination failure the ministry had been meant to solve (Misbar, 2024).
Treated separately, these two claims sit in separate literatures. The first belongs to anti-semitism studies and platform governance research: the province of scholars such as Monika Schwarz-Friesel and Monika Hübscher, and of the monitoring organizations that count and classify hateful content (Hübscher, 2023; Hübscher & von Mering, 2022; Kaati et al., 2021; Schwarz-Friesel, 2019). The second belongs to public diplomacy and strategic communication, where Israeli hasbara is analyzed as an institution adapting poorly to a networked information environment (Siman-Tov et al., 2023). Treated together, they describe a single information environment. The platforms amplifying hatred of Jews are the same platforms on which the Israeli state has failed to explain itself, and the architecture producing both outcomes is the same. Engagement-optimized recommendation systems (the algorithms that decide which posts users see according to how likely those posts are to provoke clicks, comments, and shares) do not distinguish between a conspiracy theory about Jewish control of finance and a government press release about a hostage negotiation. Both are ranked against the same metric, and the conspiracy theory generally wins, because content expressing hostility toward an out-group travels further than procedural explanation. In a study of 2.7 million Facebook and Twitter posts, each additional word referring to a political out-group raised the odds of a post being shared by 67 percent (Rathje et al., 2021), and recommendation algorithms that prioritize engagement reward such polarizing narratives accordingly (Santus & Bettin, 2025, p. 15).
This paper argues that the two literatures need to be read against one another. Digital anti-semitism cannot be fully explained by platform incentive structures alone, because those structures operate on a battlefield whose terms (what counts as anti-Semitic criticism of Israel and what counts as legitimate criticism) remain unsettled among scholars, states, and platforms themselves. Nor can the failure of Israeli public diplomacy be explained purely as a communications problem, a matter of better messaging or a more coordinated ministry, because the environment in which it operates has structural features, such as virality, anonymity, and algorithmic curation, that reward its adversaries’ content more reliably than its own, regardless of message quality. Section 2 traces the definitional dispute; Section 3 examines platform architecture; Section 4 analyzes the collapse of hasbara after October 7; Section 5 considers where the line between critique and anti-semitism can and cannot be drawn; Section 6 discusses October 7 as an inflection point; and Section 7 asks what the European Union’s regulatory response can and cannot accomplish while the underlying dispute over definitions remains open.
2. The Definitional Contest: Anti-semitism, “New Anti-semitism,” and the IHRA Working Definition
No serious account of digital anti-semitism can proceed without confronting a definitional dispute that predates social media by decades and that social media has accelerated rather than resolved. The International Holocaust Remembrance Alliance (IHRA), an intergovernmental body of more than thirty member countries, adopted a non-legally binding “working definition” of anti-semitism in May 2016. The definition describes anti-semitism as “a certain perception of Jews, which may be expressed as hatred toward Jews,” and is accompanied by eleven illustrative examples, seven of which concern the State of Israel (Behar, 2021; International Holocaust Remembrance Alliance, 2016; World Jewish Congress, n.d.). The document also states that “criticism of Israel similar to that leveled against any other country cannot be regarded as anti-Semitic,” a clause its defenders cite as evidence that it does not target ordinary political disagreement (Porat et al., 2021).
Its critics read the same document differently. Kenneth Stern, who led the drafting of the 2005 text from which the IHRA definition derives, has argued that it is being “weaponized” to suppress advocacy for Palestinian rights. This criticism carries particular weight given his authorship (Stern, 2019). A 2020 joint statement by Palestinian and Arab academics, journalists, and intellectuals argued that the definition’s examples conflate Judaism with Zionism, constructing a false equivalence between Jewish identity and support for a specific state’s policies (“Palestinian Rights and the IHRA Definition,” 2020). The Center for Security, Race and Rights at Rutgers Law School has gone further, describing the definition as a politicized legal tool whose primary function is not to protect Jews but to shield Israeli policies, including occupation and settlement expansion, from criticism and accountability (Center for Security, Race and Rights, 2023).
This dispute is not new, and treating it as a product of October 7 or of social media misreads its history. The claim that hostility toward Israel constitutes a novel form of Jew-hatred, the so-called “new anti-semitism,” was popularized in the 1970s (Forster & Epstein, 1974). Brian Klug’s long-running critique of that thesis argued as early as 2003 that the concept was analytically confused, resting on an equivalence between the individual Jew of classical anti-semitism and the State of Israel that does not hold up under scrutiny (Klug, 2003, 2013). Jonathan Judaken’s 2008 survey disentangled the several distinct strands that the label “new anti-semitism” had bundled together. It noted that information technologies had both facilitated the global spread of anti-semitism and furnished new means of combating it, an early sign that the definitional dispute and the digital one would converge (Judaken, 2008). Judaken’s more recent work extends his critique, arguing against claims that Judeophobia is unique and showing how it is entangled with other racisms, including Islamophobia and xenophobia (Judaken, 2024).
The stakes of this argument are not merely academic. Waxman, Schraub, and Hosein (2022) identify four distinct approaches to determining whether a statement or act is anti-Semitic and show that, because these approaches frequently reach different verdicts on the same conduct, disagreement over allegations of anti-semitism, especially allegations concerning Israel and anti-Zionism, is built into the debate rather than incidental to it. A platform content moderator, an EU regulator implementing the Digital Services Act, or an algorithm trained to flag hate speech must, in practice, resolve a question that specialists have not resolved in decades of debate. This is the structural problem beneath everything that follows: digital anti-semitism research often proceeds as though the category it studies were stable, when the scholarly literature it draws on treats that category as one of the most contested in the field.
For this paper, the position taken is neither that the IHRA definition should be discarded nor that it should be treated as beyond critique, but that its contested status is itself a variable that must enter any analysis of digital anti-semitism, because it determines what a platform, a government, or a monitoring organization counts when it reports a spike in anti-Semitic content. When ISD or CyberWell report increases of fifty-fold or 86 percent, the reported number is only as stable as the definition used to generate it. As Waxman, Schraub, and Hosein (2022) show, different approaches to identifying anti-Semitism often reach different verdicts on the same conduct; researchers applying different definitions to the same body of posts would likely produce different counts.
3. Platforms as Amplifiers: Algorithmic Architecture and the Post-October 7 Surge
Whatever position one takes on the definitional dispute, platform-level volume data are harder to contest than classification data. They point to a consistent structural finding: recommendation systems optimized for engagement amplify emotionally charged and polarizing content regardless of its factual basis, and anti-Semitic narratives, which cast Jews as a hostile and conspiratorial out-group, express the kind of out-group animosity that engagement-driven platforms have been shown to reward (Rathje et al., 2021). Monika Hübscher’s concept of algorithmic anti-semitism captures this directly, arguing that platform architecture, not merely user behavior, shapes the circulation of anti-Semitic material through recommendation and curation systems designed to maximize engagement (Hübscher, 2023; see also Santus & Bettin, 2025). This is a meaningful departure from earlier accounts of online hate that treated platforms as neutral conduits for user-generated content. On this view, the architecture itself rewards divisive material, creating conditions favorable to anti-Semitic content independent of any single user’s intent.
The empirical record since October 7 supports this. ISD’s YouTube analysis found the volume of anti-Semitic comments rising more than fifty-fold in the three days following the Hamas attack; the same study recorded more than a three-fold increase in anti-Semitic keywords on fringe platforms such as 4chan, BitChute, Gab, and Telegram (Rose et al., 2023). ISD chose YouTube in part because its content resembles that of X, which allowed the researchers to reuse automated classifiers (software models that sort text into categories) developed in earlier work on X. The researchers noted that data-access restrictions imposed by X, Meta, and TikTok prevented comparable analysis of those platforms. As a result, the picture across the full range of major platforms is likely incomplete (Rose et al., 2023). CyberWell, an Israeli nonprofit that serves as a “trusted flagger” for Meta and TikTok (an organization whose reports of policy-violating content the platforms review on a priority basis), documented an 86 percent increase in online anti-semitism in the three weeks following the attack. It also identified content denying or minimizing the October 7 massacre, particularly the sexual violence committed by Hamas, as a distinct and persistent category of concern (CyberWell, 2024; Santus & Bettin, 2025, p. 11). CCDH’s platform-specific findings, that X continued hosting 85 percent of reported anti-Semitic posts a week after they were flagged and that Holocaust denial content remained easy to find on TikTok despite stated policy, point to enforcement failure compounding the amplification problem rather than substituting for it (Lavelle, 2023; Navarro, 2024).
Two features of this architecture deserve separate attention because they distinguish the current moment from earlier phases of online anti-semitism research. The first is linguistic sophistication as an adaptation to moderation rather than a retreat from it. The Decoding Anti-semitism research project, which has analyzed large corpora of online comments across several European countries, finds that users “resort to regularly updated discursive strategies, such as wordplay, allusions and coded memes” precisely “to avoid detection by human or automated moderators” (Becker et al., 2024, p. 3). Santus and Bettin (2025, p. 9), drawing on Miehling’s (2024) taxonomy of online anti-Semitic hate speech, describe the result as “semantic camouflage”: euphemism, coded language, and implicit references that evade keyword-based detection while remaining fully legible to the intended audience. This is not merely evasion; it is co-evolution with the moderation systems built to contain it. It also means that raw volume counts likely understate the true prevalence of anti-Semitic content, because they miss material whose surface form has been adapted specifically to survive automated review.
The second feature is the emergence of generative artificial intelligence (AI). These systems produce new text, images, audio, or video on request, as a multiplier of production rather than merely a channel of distribution. Earlier research on digital anti-semitism in the era of “Web 2.0” (the participatory internet of blogs, social networks, and user-generated content), with Schwarz-Friesel’s work on “Anti-semitism 2.0” foremost among it, analyzed human-authored content amplified by algorithmic curation (Schwarz-Friesel, 2019). More recent scholarship documents anti-Semitic material that is itself synthetically produced, including fabricated Holocaust survivor testimony, synthetic archival images, and AI-generated content presented as geopolitical analysis or pseudo-journalism (Santus & Bettin, 2025, p. 10). Santus and Bettin argue that such content carries “a particular form of persuasive power derived from its apparent objectivity” (2025, p. 10): audiences may read the lack of an identifiable human position as evidence of impartiality rather than of automation. This is a new analytic problem, not a continuation of the Web 2.0 pattern under a new label, because it unsettles the intent-based frameworks on which much anti-semitism scholarship relies: a generative system holds no animus, yet it can reproduce and scale content that functions identically to content produced by someone who does.
Neither point should be read as an excuse for the platforms on the grounds of technical complexity. Engagement-based ranking is a design choice, not a natural feature of networked communication, and the persistence of flagged anti-Semitic content well after it has been formally reported, as CCDH documented on X, indicates a governance failure that predates, and is independent of, the generative AI question. An architecture built to maximize attention has proven equally capable of amplifying accurate reporting on Israeli military conduct, false claims about Israeli atrocities that do not withstand verification, genuine anti-Semitic conspiracy content, and Israeli government messaging, sorting among them not by truth value but by emotional charge. This is the structural fact on which the following section builds: Israel’s own communication after October 7 was graded on the same curve as the anti-Semitic content it competed against, and lost.
4. Hasbara 2.0: The Israeli State’s Digital Narrative Apparatus
Hasbara, the Hebrew term conventionally translated as “explanation,” describes Israel’s long-standing approach to public diplomacy, one whose roots predate the state itself in the efforts of early Zionist figures such as Nahum Sokolow to explain the movement’s aims to skeptical audiences (Misbar, 2024). Its digital iteration, which Miriyam Aouragh (2016) terms “hasbara 2.0,” took shape after the 2008–2009 Gaza war, when the Israel Defense Forces launched a blog, a Twitter account, and a YouTube channel, and it was progressively professionalized through the Gaza operations of 2012 and 2014. Its premise remained explanatory: traditional tools such as press releases, formal communiqués, and embassy briefings could be carried onto social media largely intact. Recent Israeli analysis argues that this premise, presenting an accurate account of events to explain and educate, is precisely what has failed in a media environment built around emotional engagement (Siman-Tov et al., 2023).
The failure of this adaptation after October 7 is not a claim advanced only by critics of Israeli policy; Israeli analysts and sympathetic observers abroad widely acknowledge it. Ilan Berman of the American Foreign Policy Council writes bluntly that Israel has long struggled to convey its point of view and that the months after October 7 were a “wake-up call,” revealing a state unprepared for the global surge of anti-Israel and anti-Semitic sentiment that followed (Berman, 2024). Writing in the Jerusalem Post in November 2025, David Siman-Tov, a senior researcher at Israel’s Institute for National Security Studies (INSS), describes hasbara’s core assumption- that sufficiently clear explanation will produce international understanding and support- as “naïve” in an era when “social media algorithms decide what appears credible – and what is buried,” and argues that Israel “excels on the battlefield but falters in the battle for hearts and minds” (Siman-Tov, 2025).
Scholarly analysis of this dynamic identifies an asymmetry in how narratives are constructed, not merely a gap in resources. Siman-Tov, Michlin-Shapir, and Fridman argue that traditional hasbara seeks to explain and educate by presenting an accurate account of events. In contrast, Hamas embeds its messaging in the broader Palestinian story with which audiences already identify, appealing to emotion and adjusting its narrative to what its target audiences want to hear (Siman-Tov et al., 2023). This finding aligns with the platform architecture discussed above. On any platform whose ranking systems reward engagement over accuracy, content built to inform competes at a structural disadvantage against content built to provoke, and the disadvantage holds regardless of the substantive merit of the informational content.
The institutional response compounded rather than resolved the problem. The closure of Israel’s Ministry of Public Diplomacy in October 2023, in the first weeks of the war, removed a coordinating body at precisely the moment when coordination across the Ministry of Foreign Affairs, the military spokesperson’s office, the Ministry of Tourism, and the Prime Minister’s Office was most needed. Commentary on the period has identified the resulting lack of unified messaging as a central weakness, reporting that within weeks of the attack no government representative was coordinating Israel’s public diplomacy efforts (Jerusalem Post, 2023; Misbar, 2024). More recent Israeli commentary calls for abandoning the hasbara framework altogether in favor of what is termed information warfare. Jeffrey Kahn argues that the change required goes beyond terminology (“naming shapes posture, posture shapes strategy, and strategy shapes outcomes”) and calls for “the integration of intelligence, operational planning, and strategic communications,” a shift from reactive explanation toward continuous, proactive management of Israel’s global image (Kahn, 2026).
This reframing carries its own risks, which the history of hasbara should make legible. In a 2012 address published by the Middle East Policy Council, Chas W. Freeman Jr., a former U.S. ambassador to Saudi Arabia, described hasbara as reaching well beyond public relations into an instrument that seeks to “determine the way issues are defined by the media, the intelligentsia, and social networks,” to “establish the parameters of politically correct discourse,” and to “delegitimize both critics and their arguments” (Freeman, 2012). A shift from hasbara to explicit information warfare does not resolve the tension between explaining a position and managing its perception; it names that tension more candidly. That may improve Israel’s competitive position on social media while doing nothing to settle whether the content being amplified is accurate, and nothing to resolve the definitional dispute, addressed in Section 2, over what counts as legitimate criticism and what counts as anti-Semitic delegitimization. A narrative apparatus optimized for the same engagement metrics that reward anti-Semitic conspiracy content will, if successful, itself become optimized for emotional intensity over documentary precision, converging with its adversary’s method even as it opposes its adversary’s substance.
5. Between Critique and Contempt: Locating the Analytic Line
The preceding two sections describe two outcomes of a single architecture: anti-Semitic content is amplified because it is emotionally potent, and Israeli state messaging is not amplified in comparable measure, for the same structural reason applied in reverse. This structural parallel should not be mistaken for moral equivalence, and the literature reviewed in Section 2 exists precisely because scholars have not agreed on where to draw the line between hateful content and legitimate political speech.
Several empirical patterns from the post-October 7 period show where the line is contested rather than clear. Holocaust inversion, the rhetorical move in which “the Israelis are cast as the ‘new’ Nazis and the Palestinians as the ‘new’ Jews” (Klaff, 2014), has a documented history predating the current war by more than a decade and has been treated by scholars of contemporary anti-semitism as a recognizable and analyzable pattern (Klaff, 2014; Santus & Bettin, 2025, p. 13). The move does not merely criticize a specific policy; it collapses a historically specific atrocity into a template applied to a contemporary conflict, flattening both events. Even this comparatively well-defined pattern, however, sits on the fault line between definitions. The IHRA definition lists “drawing comparisons of contemporary Israeli policy to that of the Nazis” among its examples of anti-semitism (International Holocaust Remembrance Alliance, 2016). The Jerusalem Declaration on Anti-semitism, the principal alternative drafted in 2021 by scholars critical of the IHRA text, does not list Nazi comparisons among its examples. However, it does classify as anti-Semitic “applying the symbols, images and negative stereotypes of classical anti-semitism” to the State of Israel (Jerusalem Declaration on Anti-semitism, 2021). Under the IHRA framework, then, Holocaust inversion is anti-Semitic as such; under the Jerusalem Declaration, it is anti-Semitic when it carries classical anti-Semitic imagery or stereotypes, and otherwise falls to case-by-case judgment.
Other patterns are considerably harder to classify, and candor about this difficulty is part of the analytic contribution this paper aims to make. Criticism of Israeli military conduct in Gaza, including the use of the term genocide to characterize that conduct, occupies contested ground. Critics of the IHRA definition argue that such criticism has been improperly folded into accusations of anti-semitism; the Jerusalem Declaration, for instance, holds that evidence-based criticism of Israel as a state, including comparisons with other historical cases such as settler colonialism or apartheid, is not in and of itself anti-Semitic (Center for Security, Race and Rights, 2023; Jerusalem Declaration on Anti-semitism, 2021). Others argue that genocide accusations against Israel frequently carry older anti-Semitic tropes. The historian Norman Goda, for example, contends that such accusations have been leveled at Israel continuously since its founding and are often joined to claims that associate all Jews with the alleged crime or attribute to Israel special powers to manipulate Western governments (Goda, 2025). Both positions can be true at once without collapsing into each other. A given social media post alleging genocide may reflect a serious, empirically contestable claim about proportionality and civilian harm, or it may function as a vehicle for collective blame attached to Jews as a category; platform-scale content moderation, whether human or algorithmic, is poorly equipped to make this distinction at the volume required.
This is where the definitional dispute traced in Section 2 becomes operationally consequential rather than merely academic. A platform, a regulator, or an AI-driven classifier trained on the IHRA definition and its examples will flag a meaningfully different set of posts than one trained on Klug’s framework or on the approach favored by the signatories of the 2020 Palestinian and Arab statement discussed above. The resulting figures (the fifty-fold and 86 percent increases reported by ISD and CyberWell, and the enforcement failures documented by CCDH) are real and are corroborated across monitoring organizations using different methodologies, which lends confidence to the finding of a genuine surge in hostile content. But the composition of that surge remains an open empirical and normative question that volume metrics alone cannot answer: how much of it reflects classical anti-Semitic tropes repackaged for a new conflict, how much reflects contested but good-faith political criticism, and how much occupies the ambiguous space between the two.
6. Discussion: October 7 as an Inflection Point Rather Than an Origin
It is worth resisting the framing, common in both journalistic and some scholarly accounts, that October 7 created the dynamics described above. The more defensible reading, consistent with the literature reviewed here, is that October 7 functioned as what Gruber, Loy, and Poensgen (2024) call an opportunity structure: a set of social, political, and situational conditions that does not cause anti-semitism but gives people already holding anti-Semitic views a framework in which to act on them. On this reading, the attack activated and rendered visible anti-Semitic sentiment and networks that predated it by years, rather than generating a new phenomenon (Gruber et al., 2024; Santus & Bettin, 2025, p. 11). The same is true of Israel’s public diplomacy weaknesses: hasbara’s structural mismatch with social media’s emotional economy predates October 7 by more than a decade, and the war exposed rather than created that mismatch (Aouragh, 2016; Reut Institute, 2010).
This distinction matters for policy because it implies different remedies from those a crisis-response framing would suggest. If October 7 created the problem, the appropriate response is crisis management: better rapid-response communication, faster content moderation during acute events, and temporary surge capacity at platforms and monitoring organizations. If October 7 exposed a pre-existing structural problem, the appropriate response is architectural: addressing, as a general matter and independent of any single conflict, the recommendation systems that reward emotional intensity over accuracy, and confronting the definitional dispute that determines what those systems should be trained to flag in the first place. The evidence reviewed here supports the latter reading considerably more than the former. Algorithmic anti-Semitism, in Hübscher’s formulation, describes a standing feature of platform design, not an emergency condition triggered by a single attack, and the shortcomings of hasbara as a communications model were diagnosed well before October 7, not discovered in its aftermath. In 2010, the Israeli think tank the Reut Institute warned that public relations, “while critically important,” “cannot and will not neutralize” campaigns to delegitimize Israel, which operate from the bottom up through grassroots movements, academia, and social networks (Reut Institute, 2010). In 2016, Israel’s State Comptroller found that a lack of cooperation between the Foreign Ministry and the military spokesperson had produced “an advocacy failure” during the 2014 Gaza war, and that the Foreign and Strategic Affairs ministries were locked in “an active power struggle” over responsibility for countering boycott campaigns (Wootliff, 2016).
This has an uncomfortable implication for advocates on every side of the underlying conflict. If the platform architecture rewards emotional intensity regardless of factual accuracy, then improving the accuracy of one’s own side’s content, whether through more rigorous documentation of Hamas’s October 7 atrocities or more rigorous documentation of civilian harm in Gaza, will not by itself improve that content’s circulation relative to less accurate but more emotionally potent competing narratives. Better facts, delivered through the same engagement-optimized channels, do not reliably outcompete worse facts wrapped in a more affecting narrative. This is a harder problem than either “fix the definition” or “fix the platforms” acknowledges on its own. Even a perfectly resolved definitional dispute, cleanly separating anti-Semitic content from legitimate criticism, would still leave the accurate version of legitimate criticism and the anti-Semitic version competing on unequal terms, sorted by an algorithm indifferent to which is which.
7. Regulatory Horizon: The Digital Services Act and the Limits of Platform Governance
The European Union’s Digital Services Act (DSA) entered into force in November 2022, began applying to the largest platforms in August 2023, and became fully applicable in February 2024. It requires Very Large Online Platforms (VLOPs), those with more than 45 million monthly users in the EU, to act against illegal content, including hate speech that is unlawful under EU or national law, and to assess and mitigate the systemic risks their services create (Regulation (EU) 2022/2065, 2022). It thus represents the first significant regulatory attempt to hold platforms accountable for the architecture discussed throughout this paper rather than merely for individual pieces of content. ISD’s YouTube research was explicitly framed around this framework, treating the DSA’s arrival as an occasion to “evaluate platform responsibilities” as a VLOP (Rose et al., 2023).
The regulation’s limits track the definitional problem this paper has emphasized throughout. The DSA obliges platforms to act against illegal content but leaves the definition of illegality to EU and national law. EU law sets only a minimum standard: Council Framework Decision 2008/913/JHA requires member states to criminalize public incitement to violence or hatred on grounds including religion and ethnic origin, and the public condoning, denial, or gross trivialization of genocide and related crimes, including the Holocaust. Beyond that floor, national laws differ considerably, and transposition of the Framework Decision has been uneven (Council of the European Union, 2008; European Commission, 2021). A post that a German court would treat as incitement to hatred (Volksverhetzung) under Section 130 of the German Criminal Code may be lawful in another member state. A platform operating across the European market must therefore either restrict access to such content only in the member states where it is unlawful, leaving the same post visible elsewhere in the Union, or apply a single standard across all member states, in which case the platform itself, rather than any legislature, decides where the line falls. Legal analysis of the IHRA definition points to a related difficulty: a legal opinion commissioned by the Swiss government concluded that the definition’s examples lack the precision required of a legal standard and depend heavily on contextual interpretation (Hertig Randall & Imbeck, 2020).
This is not an argument against the DSA or against platform regulation generally. The enforcement failures CCDH documented, with content persisting well after formal flagging, indicate real governance gaps that regulation can plausibly narrow regardless of definitional disputes. Requiring platforms to act promptly on flagged content, to publish transparency reports on their moderation decisions, and to submit to independent audits addresses the enforcement gap directly without requiring prior resolution of the definitional question. But it is an argument against expecting regulation to resolve what this paper has argued is the prior and more fundamental problem. A platform can enforce a definition faster and more consistently without that definition commanding scholarly consensus, and the resulting enforcement, however improved, will inherit whatever contestation attaches to the definition being enforced. The IHRA definition’s adoption or endorsement by a growing number of governments, including seventeen EU member states by 2021, and its status as a benchmark in the European Commission’s strategy on combating anti-semitism give platforms a common reference point that did not previously exist in codified form, which is a real institutional achievement (European Commission, 2021). It does not, and given the nature of the underlying scholarly dispute, cannot, settle whether that reference point draws the line between criticism and hatred where most scholars of anti-semitism would independently choose to draw it.
8. Conclusion
The relationship between digital anti-semitism and the Israeli narrative in social media environments is not one of separate problems that happen to share a news cycle. It is one problem with two faces, produced by an architecture that grades all content (hateful and legitimate, accurate and false, Israeli and Palestinian, Jewish and anti-Semitic) against a single metric of emotional engagement, with no mechanism of its own for distinguishing among these categories. Anti-Semitic content surged after October 7 because that architecture rewards the dehumanization and conspiracy narratives such content typically carries; Israeli public diplomacy collapsed over the same period because hasbara, built around explanation rather than emotional appeal, was competing on the same terms and losing by design rather than by accident. Resolving the definitional dispute over where legitimate criticism of Israel ends and anti-Semitism begins would clarify what platforms and regulators are meant to enforce. Still, it would not, by itself, repair the architecture that determines what gets amplified, whichever side of that line a given post falls on. The scholarship reviewed here has advanced considerably in documenting volume and in theorizing algorithmic amplification. It has advanced far less in reckoning with the fact that the definitional question underlying all of it remains as unsettled now as it was when Klug challenged the “new anti-semitism” thesis in 2003, a year before Facebook was founded.
About the author: Aaron Walter, MBA, Ph.D., Mykolas Romeris University, Vilnius; Masaryk University, Brno (Adjunct)
Sources: Walter, Aaron T. 2026. “Digital Antisemitism: Israeli Narrative After October 7.” Paper presented at the 14th Annual Conference of the European Association of Israel Studies (EAIS), Vytautas Magnus University, Kaunas, Lithuania, October 11–13.
American Jewish Committee. (2024, February 18). Anti-semitism online and on social media runs rampant. https://www.ajc.org/news/antisemitism-online-and-on-social-media-runs-rampant
Aouragh, M. (2016). Hasbara 2.0: Israel’s public diplomacy in the digital age. Middle East Critique, 25(3), 271–297. https://doi.org/10.1080/19436149.2016.1179432
Becker, M. J., Ascone, L., Placzynta, K., & Vincent, C. (Eds.). (2024). Anti-semitism in online communication: Transdisciplinary approaches to hate speech in the twenty-first century. Open Book Publishers. https://doi.org/10.11647/OBP.0406
Behar, M. (2021, March 25; updated 2025, March 14). The IHRA’s careless conflations on anti-semitism (and a few alternatives). Contending Modernities, University of Notre Dame. https://contendingmodernities.nd.edu/global-currents/inhra-careless-conflations/
Berman, I. I. (2024, April 24). Hasbara doesn’t work: Israel needs a new form of messaging. American Foreign Policy Council. https://www.afpc.org/publications/articles/hasbara-doesnt-work-israel-needs-a-new-form-of-messaging
Center for Security, Race and Rights. (2023). Presumptively antisemitic: Islamophobic tropes in the Palestine–Israel discourse. Rutgers Law School. https://csrr.rutgers.edu/issues/presumptively-antisemitic/
Council of the European Union. (2008). Council Framework Decision 2008/913/JHA of 28 November 2008 on combating certain forms and expressions of racism and xenophobia by means of criminal law. Official Journal of the European Union, L 328, 55–58. https://eur-lex.europa.eu/eli/dec_framw/2008/913/oj
CyberWell. (2024, October). The evolution of online anti-semitism: Pre- & post-October 7. https://cyberwell.org/reports/the-evolution-of-online-antisemitism-pre-post-october-7/
European Commission. (2021). EU strategy on combating anti-semitism and fostering Jewish life (2021–2030) (COM(2021) 615 final). https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52021DC0615
Forster, A., & Epstein, B. R. (1974). The new anti-Semitism. McGraw-Hill.
Freeman, C. W., Jr. (2012, December 3). Hasbara and the control of narrative as an element of strategy [Speech]. Middle East Policy Council. https://mepc.org/speeches/hasbara-and-control-narrative-element-strategy/
Goda, N. J. W. (2025). The genocide libel: How the world has charged Israel with genocide (ISCA Research Paper 2025-3). Institute for the Study of Contemporary Anti-semitism, Indiana University Bloomington. https://isca.indiana.edu/publication-research/research-paper-series/norman-jw-goda-research-paper.html
Gruber, J., Loy, B., & Poensgen, D. (2024). Antisemitic reactions to October 7: The German case. In R. Freedman & D. Hirsh (Eds.), Responses to 7 October: Law and society (pp. 91–100). Routledge. https://doi.org/10.4324/9781003497417-11
Hertig Randall, M., & Imbeck, C. (2020, November 6). The IHRA working definition of anti-semitism: A legal analysis [Legal opinion commissioned by the Service for Combating Racism, Swiss Federal Department of Home Affairs]. https://archive-ouverte.unige.ch/unige:152206
Hübscher, M. (2023). Algorithmic anti-semitism on social media. In M. Weitzman, R. J. Williams, & J. Wald (Eds.), The Routledge history of anti-semitism (pp. 364–372). Routledge. https://doi.org/10.4324/9780429428616-40
Hübscher, M., & von Mering, S. (Eds.). (2022). Anti-semitism on social media. Routledge. https://doi.org/10.4324/9781003200499
International Holocaust Remembrance Alliance. (2016). Working definition of anti-semitism. https://holocaustremembrance.com/resources/working-definition-antisemitism
Jerusalem Declaration on Anti-semitism. (2021). The Jerusalem Declaration on Anti-semitism. https://jerusalemdeclaration.org/
Jerusalem Post. (2023, October 22). Israel shuts down public diplomacy ministry, budget heads to South. https://www.jpost.com/israel-news/politics-and-diplomacy/article-769602
Judaken, J. (2008). So what’s new? Rethinking the ‘new anti-semitism’ in a global age. Patterns of Prejudice, 42(4–5), 531–560. https://doi.org/10.1080/00313220802377453
Judaken, J. (2024). Critical theories of anti-Semitism. Columbia University Press. https://doi.org/10.7312/juda21292
Kaati, L., Asplund Cohen, K., & Pelzer, B. (2021). Anti-semitism in social media: Conspiracies, stereotypes, and Holocaust denial (FOI-R--5198--SE). Swedish Defence Research Agency (FOI). https://www.foi.se/en/foi/reports/report-summary.html?reportNo=FOI-R--5198--SE
Kahn, J. (2026, January 11). What’s in a name? A lot: Why Israel must reframe hasbara as informational warfare. The Jerusalem Post. https://www.jpost.com/opinion/article-883000
Klaff, L. (2014). Holocaust inversion and contemporary anti-semitism. Fathom, Winter 2014. https://fathomjournal.org/holocaust-inversion-and-contemporary-antisemitism/
Klug, B. (2003). The collective Jew: Israel and the new anti-semitism. Patterns of Prejudice, 37(2), 117–138. https://doi.org/10.1080/0031322032000087973
Klug, B. (2013). Interrogating ‘new anti-semitism’. Ethnic and Racial Studies, 36(3), 468–482. https://doi.org/10.1080/01419870.2013.734385
Lavelle, T. (2023, November 6). Fact-checking TikTok’s claims on anti-semitism. Center for Countering Digital Hate. https://counterhate.com/blog/fact-checking-tiktoks-claims-on-antisemitism/
Misbar. (2024, October 25). Israel’s global challenge in public diplomacy and ‘hasbara’ following October 7. https://www.misbar.com/en/editorial/2024/10/25/israels-global-challenge-in-public-diplomacy-and-lsquohasbararsquo-following-october-7
Navarro, M. (2024, February 23). Understanding anti-semitism on social media. Center for Countering Digital Hate. https://counterhate.com/blog/understanding-antisemitism-on-social-media/
Palestinian rights and the IHRA definition of anti-semitism. (2020, November 29). The Guardian. https://www.theguardian.com/news/2020/nov/29/palestinian-rights-and-the-ihra-definition-of-antisemitism
Porat, D., Quer, G., & Naamat, T. (2021, December 23). The IHRA working definition of anti-semitism: Criticism, implementation, and importance. Institute for National Security Studies. https://www.inss.org.il/publication/ihra/
Rathje, S., Van Bavel, J. J., & van der Linden, S. (2021). Out-group animosity drives engagement on social media. Proceedings of the National Academy of Sciences, 118(26), Article e2024292118. https://doi.org/10.1073/pnas.2024292118
Regulation (EU) 2022/2065 of the European Parliament and of the Council of 19 October 2022 on a Single Market for Digital Services and amending Directive 2000/31/EC (Digital Services Act). (2022). Official Journal of the European Union, L 277, 1–102. https://eur-lex.europa.eu/eli/reg/2022/2065/oj
Reut Institute. (2010). Building a political firewall against Israel’s delegitimization: Conceptual framework. https://www.reutgroup.org/Publications/building-a-political-firewall--against-israel%27s-delegitimization-
Rose, H., Comerford, M., & Guhl, J. (2023, October 31; updated 2024, February 7). Rise in anti-semitism on both mainstream and fringe social media platforms following Hamas’ terrorist attack. Institute for Strategic Dialogue. https://www.isdglobal.org/digital-dispatch/rise-in-antisemitism-on-both-mainstream-and-fringe-social-media-platforms-following-hamas-terrorist-attack/
Santus, D., & Bettin, C. (2025). Anti-semitism in the algorithmic age: Propaganda, disinformation, and generative AI. Zeitschrift für Religion, Gesellschaft und Politik. Advance online publication. https://doi.org/10.1007/s41682-025-00234-6
Schwarz-Friesel, M. (2019). Judenhass im Internet: Antisemitismus als kulturelle Konstante und kollektives Gefühl [Jew-hatred on the internet: Antisemitism as a cultural constant and collective emotion]. Hentrich & Hentrich. https://www.hentrichhentrich.de/buch-judenhass-im-internet.html
Siman-Tov, D. (2025, November 19). Israel’s fight for influence in the digital age. The Jerusalem Post. https://www.jpost.com/jerusalem-report/article-874419
Siman-Tov, D., Michlin-Shapir, V., & Fridman, O. (2023, November 8). From hasbara (public diplomacy) to influence in the Gaza war (INSS Insight No. 1783). Institute for National Security Studies. https://www.inss.org.il/publication/hasbara/
Stern, K. S. (2019, December 13). I drafted the definition of anti-semitism. Right-wing Jews are weaponizing it. The Guardian. https://www.theguardian.com/commentisfree/2019/dec/13/antisemitism-executive-order-trump-chilling-effect
Waxman, D., Schraub, D., & Hosein, A. (2022). Arguing about anti-semitism: Why we disagree about anti-semitism, and what we can do about it. Ethnic and Racial Studies, 45(9), 1803–1824. https://doi.org/10.1080/01419870.2021.1960407
Wootliff, R. (2016, May 24). Dysfunctional Israeli leadership failing utterly in battle against BDS, state watchdog warns. The Times of Israel. https://www.timesofisrael.com/dysfunctional-israeli-leadership-failing-utterly-in-battle-against-bds-state-watchdog-warns/
World Jewish Congress. (n.d.). Misconception: IHRA conflates criticism of Israel with anti-semitism. Retrieved September 25, 2026, from https://www.worldjewishcongress.org/en/ihra-misconceptions/Misconception-IHRA-conflates-criticism-of-Israel-with-antisemitism