---
title: "Incident Debrief: Ripple Effects of the Charlie Kirk Assassination in the Canadian Information Ecosystem"
authors: aengus-bridgman, eric-merkley, mathieu-lavigne, chris-ross, mika-desblancs-patel, esli-chan, ben-steel, diya-jiang, zeynep-pehlivan, marcus-kolga, layla-mashkoor, meredith-furbish
date: 2026-01-28
doi: 10.66536/charlie-kirk-debrief_2026
pdf: /files/publications/charlie-kirk-debrief_2026.pdf
permalink: https://mediatechdemocracy.com/en/publications/charlie-kirk-debrief_2026/
---

**Ripple Effects of the Charlie Kirk Assassination in the Canadian Information Ecosystem** 

January 2026

# **Incident Debrief**

# **Ripple Effects of the Charlie Kirk Assassination in the Canadian Information Ecosystem**

January 2026

### **INCIDENT RESPONSE TEAM**

This incident response notification is issued by the Canadian Digital Media Research Network (CDMRN). Organizations contributing to this response include: [Disinfowatch](https://disinfowatch.org/), [the DFRLab](https://dfrlab.org/), and the [Media Ecosystem Observatory](https://www.mediaecosystemobservatory.com/).

### **KEY CONTACTS**

The following is a list subject matter experts involved in the investigation that are available for contact:

### **Aengus Bridgman,**

*Director, Media Ecosystem Observatory*

### [aengus.bridgman@mcgill.ca](mailto:aengus.bridgman@mcgill.ca)

Expertise: Mis- and disinformation, social media network analysis, narrative analysis

### **Marcus Kolga**

Director, DisinfoWatch [marcuskolga@gmail.com](mailto:marcuskolga@gmail.com) Expertise: Central and Eastern European politics; disinformation and propaganda; media and foreign policy

### **Layla Mashkoor**

Deputy Director of Research, Digital Forensic Research Lab [lmashkoor@atlanticcouncil.org](mailto:lmashkoor@atlanticcouncil.org) Expertise: Foreign interference, information manipulation, social media analysis, tech policy

### **Isabelle Corriveau**

Associate Director, Public engagement [isabelle.corriveau2@mcgill.ca](mailto:isabelle.corriveau2%40mcgill.ca?subject=)

**January 2026**

### **Incident Debrief**

# **Ripple Effects of the Charlie Kirk Assassination in the Canadian Information Ecosystem**

### **Summary**

- E **Nearly all Canadians heard about the assassination** which transformed Kirk from a somewhat known online figure (~25% of Canadians) into a household name (88%), with social media serving as the primary channel for exposure and debate, particularly among young Canadians.
- E **Online discourse was intense but short-lived** with initial conspiracy theories and speculation giving way to polarized blame and ideological contestation before subsiding within roughly two weeks. The most extreme claims were challenged by other social users.
- E **Online reactions sharply diverged from offline ones**; while social media amplified anger, blame, and conspiratorial narratives, the most common reactions among Canadians were sadness, sympathy, and indifference. These emotions were relatively muted in the online conversation where anger, disgust, outrage, relief, and happiness were disproportionately present.
- E **Russian state-affiliated actors exploited the event** with disproportionate coverage and heavily politicized, conspiratorial narratives. China and India did not follow suit.
- E Despite near-universal condemnation of the assassination, **concerns about political violence are chilling democratic participation**; half of Canadians say they avoid activities like posting opinions online or attending protests due to fear of violence.

On September 10, 2025, Charlie Kirk, prominent American rightwing activist and founder of Turning Point USA, was assassinated while at a public rally on the campus of Utah Valley University. His death resulted in an apparent surge of highly charged and polarized online conversations, disinformation and calls for political violence and retribution in the American but also in

**Janvier 2026**

### **Compte rendu d'incident**

# **Effets d'entraînement de l'assassinat de Charlie Kirk dans l'écosystème canadien de l'information**

## **Résumé**

- E **Presque tous les Canadiens ont entendu parler de l'assassinat**, qui a fait passer Kirk d'une figure en ligne relativement connue (par environ 25 % des Canadiens) à une figure connue par la grande majorité des Canadiens (88 %). Les réseaux sociaux ont constitué le principal canal d'exposition à cette histoire et de débat, surtout chez les jeunes Canadiens.
- E **Les échanges en ligne ont été intenses, mais de courte durée** : les premières théories du complot et spéculations ont rapidement laissé place à une polarisation des blâmes et à des affrontements idéologiques, avant de revenir plus à la normale après environ deux semaines. Les affirmations les plus extrêmes ont été contestées par d'autres utilisateurs.
- E **Les réactions en ligne ont fortement divergé des réactions hors ligne**. Les réactions les plus fréquentes parmi les Canadiens ont été la tristesse, la sympathie et l'indifférence. Ces émotions étaient relativement peu présentes dans les conversations en ligne, où la colère, le dégoût, l'indignation, le soulagement et même la satisfaction étaient surreprésentés.
- E **Des acteurs affiliés à l'État russe ont exploité l'événement**, lui accordant une couverture disproportionnée et diffusant des récits fortement politisés et conspirationnistes. La Chine et l'Inde n'ont pas adopté la même approche.
- E Malgré une condamnation quasi unanime de l'assassinat, **les inquiétudes liées à la violence politique ont un effet dissuasif sur la participation démocratique** : la moitié des Canadiens disent éviter des activités comme publier des opinions en ligne ou assister à des manifestations par crainte de violences.

Le 10 septembre 2025, Charlie Kirk, militant américain de droite et fondateur de Turning Point USA, a été assassiné lors d'un rassemblement public sur le campus de l'Université de la Vallée de the Canadian information environment. Canadian commentators were targeted and Canadian politicians from across the political spectrum weighed in, warning about the dangers of political violence and the risks of escalating ideological conflict. l'Utah. Sa mort a entraîné une hausse marquée de conversations en ligne très polarisées et chargées émotionnellement, ainsi que de désinformation et d'appels à la violence politique et aux représailles, tant dans l'environnement informationnel américain que canadien. Des commentateurs canadiens ont été ciblés, et des responsables politiques canadiens de tous les horizons se sont exprimés pour mettre en garde contre les dangers de la violence politique et les risques d'une escalade des conflits idéologiques.

### **Incident Assessment**

This **minor information incident** created a transnational information event reaching far outside the United States and deeply into the Canadian information environment. While the initial days and weeks following the assassination felt transformational for polarization, democratic discourse, and acceptability of violence and harassment, several months after the event we can fully assess its importance and impacts. This event highlights the permeability of the Canadian information environment to major U.S. political events, the polarizing and misinformation potential of events like these, and the opportunities these events create for foreign manipulation.

The event generated widespread awareness, transforming Kirk from an online figure particularly known among the youth to a household name in Canada [[IU6](#page-24-0)]. While engagement was widespread and polarized and conspiratorial narratives were widely visible online for several weeks after the event [[IU2](#page-50-0), [IU5](#page-31-0)], the real-world consequences for Canadians have been limited. The incident appears to have had four principal consequences on the Canadian information environment:

### **1. Widespread and polarized reactions**

Nearly all Canadians were exposed to news of the event, with most exposure and discussions occurring on social media [[IU6](#page-24-0)]. Canadian news outlets primarily provided informational updates, politicians showed limited engagement, while influencers played a greater role in shaping narratives and debates [[IU2](#page-50-0), [IU5](#page-31-0)]. Divisions primarily emerged from reactions to Charlie Kirk's death and disagreement over his legacy [[IU2](#page-50-0), [IU4](#page-40-0), [IU5](#page-31-0), [IU6](#page-24-0)]. Extreme emotional reactions to the event were more visible online than offline, with the most common reactions among Canadians being sadness and sympathy for the family, shock, condemnation of violence, and concern about the future of our society [[IU6](#page-24-0)].

# **Évaluation de l'incident**

Cet **incident d'information de portée limitée** a donné lieu à un événement informationnel transnational qui a largement dépassé les frontières des États-Unis et pénétré en profondeur l'écosystème canadien de l'information. Si les jours et les semaines ayant suivi l'assassinat ont pu sembler marquants, en accentuant la polarisation, en fragilisant le discours démocratique et en banalisant la violence et le harcèlement, le recul de plusieurs mois permet aujourd'hui d'en évaluer plus précisément l'importance et les effets. L'événement met en lumière la perméabilité de l'écosystème canadien de l'information aux grands événements politiques américains, le potentiel de polarisation et de désinformation associé à ce type d'événement, ainsi que les occasions qu'ils offrent à des acteurs étrangers pour des tentatives de manipulation.

L'événement a suscité beaucoup d'attention faisant passer Kirk d'une figure en ligne surtout connue des jeunes à une personnalité connue du grand public au Canada [[IU6](#page-24-0)]. Bien que la mobilisation ait été large et polarisée, et que des récits conspirationnistes aient été très visibles en ligne pendant plusieurs semaines après l'événement [[IU2](#page-50-0), [IU5](#page-31-0)], les répercussions concrètes pour les Canadiens sont demeurées limitées. L'incident semble avoir eu quatre conséquences principales sur l'écosystème canadien de l'information :

### **1. Réactions généralisées et polarisées**

Presque tous les Canadiens ont été exposés aux nouvelles liées à l'événement, la majorité de cette exposition et des discussions ayant eu lieu sur les réseaux sociaux [[IU6](#page-24-0)]. Les médias canadiens ont principalement fourni des mises à jour factuelles, les responsables politiques se sont peu engagés, tandis que les influenceurs ont joué un rôle plus important dans la structuration des récits et des débats [[IU2](#page-50-0), [IU5](#page-31-0)]. Les divisions sont surtout apparues autour des réactions à la mort de Charlie Kirk et des désaccords quant à son leg et à son rôle public [[IU2](#page-50-0), [IU4](#page-40-0), [IU5](#page-31-0), [IU6](#page-24-0)]. Les réactions émotionnelles extrêmes ont été plus visibles en ligne qu'hors ligne ; chez les Canadiens, les réactions les plus courantes ont toutefois été la tristesse et la sympathie envers la famille, le choc, la condamnation de la violence et l'inquiétude face à l'avenir de la société [[IU6](#page-24-0)].

### **2. Foreign and domestic influencer amplification of conspiratorial content**

Polarized and conspiratorial narratives emerged primarily on X and rapidly permeated the Canadian information environment [[IU5](#page-31-0)]. While Canadian influencers were not the initial drivers of these narratives, they amplified them, extending their domestic reach [[IU2](#page-50-0), [IU5](#page-31-0)]. Russian state-affiliated media devoted disproportionate attention to the event relative to their typical North American coverage, advancing overtly politicized, conspiratorial, and emotionally charged content (consistent with a recurring pattern of exploiting Western crises to deepen division and institutional distrust) [[IU1](#page-56-0), [IU8](#page-12-0)]. Notably, neither China nor India followed suit. While the conspiratorial narratives themselves were short-lived [[IU2](#page-50-0), a sizeable minority of Canadians reported awareness of or belief in them [[IU6](#page-24-0)]. Despite this turbulent online environment, overall partisan polarization in Canada remained stable (though high) throughout 2025.

### **3. Heightened concern about political violence, despite broad condemnation**

Most Canadians believe that the assassination of Charlie Kirk is reflective of a broader problem in society [[IU3](#page-45-0)]. While political violence was widely condemned internationally and in Canada [[IU4](#page-40-0)] and support for political violence remains very low overall [[IU7](#page-17-0)], a significant share of Canadians believe that political violence is a serious problem in Canada (though a significantly smaller problem than in the U.S.) and anticipate that the problem will get worse in the coming years [[IU3](#page-45-0)]. The event also amplified partisan misperceptions, with the prominent but largely unfounded claim that the left celebrated Kirk's death reaching wide audiences online [[IU5](#page-31-0), [IU6](#page-24-0)]. Such narratives matter because partisans who overestimate their opponents' support for violence are themselves more likely to endorse it [[IU7](#page-17-0)]. While support for political violence remains low overall, the event spotlighted a troubling generational divide: young Canadians aged 18–34 are significantly more likely to view violence as sometimes necessary for social change, a share that has grown since 2024 [[IU3](#page-45-0)].

### **2. Amplification de contenus conspirationnistes par des influenceurs étrangers et domestiques**

Des récits polarisés et conspirationnistes ont émergé principalement sur X et se sont rapidement diffusés dans l'écosystème canadien de l'information [[IU5](#page-31-0)]. Bien que les influenceurs canadiens n'aient pas été à l'origine de ces récits, ils les ont amplifiés, en élargissant leur portée à l'échelle nationale [[IU2](#page-50-0), [IU5](#page-31-0)]. Des médias affiliés à l'État russe ont accordé à l'événement une attention disproportionnée par rapport à leur couverture habituelle de l'Amérique du Nord, diffusant des contenus ouvertement politisés, conspirationnistes et fortement chargés sur le plan émotionnel — conformément à un schéma récurrent consistant à exploiter les crises occidentales afin d'accentuer les divisions et la défiance envers les institutions [[IU1](#page-56-0), [IU8](#page-12-0)].

À noter que ni la Chine ni l'Inde n'ont adopté une approche comparable. Si ces récits conspirationnistes ont été de courte durée [[IU2](#page-50-0)], une minorité non négligeable de Canadiens a déclaré en avoir eu connaissance ou y avoir accordé du crédit [[IU6](#page-24-0)]. Malgré ce contexte en ligne turbulent, la polarisation partisane globale au Canada est demeurée stable — bien qu'élevée — tout au long de l'année 2025.

### **3. La violence politique suscite des inquiétudes croissantes, malgré une condamnation largement partagée**

La plupart des Canadiens estiment que l'assassinat de Charlie Kirk reflète un problème plus large au sein de la société [[IU3](#page-45-0)]. Bien que la violence politique ait été largement condamnée, tant à l'international qu'au Canada [[IU4](#page-40-0)], et que le soutien à ce type de violence demeure globalement très faible [[IU7](#page-17-0)], une part importante des Canadiens considère néanmoins que la violence politique constitue un problème sérieux au Canada — quoique nettement moins grave qu'aux États-Unis — et s'attend à ce que la situation se détériore au cours des prochaines années [[IU3](#page-45-0)]. L'événement a également accentué certaines perceptions partisanes erronées, notamment l'affirmation, très médiatisée mais largement infondée, selon laquelle la gauche aurait célébré la mort de Kirk, un récit qui a touché de vastes publics en ligne [[IU5](#page-31-0), [IU6](#page-24-0)]. Ces dynamiques sont préoccupantes, car les partisans qui surestiment le soutien de leurs adversaires à la violence sont eux-mêmes plus susceptibles de l'approuver [[IU7](#page-17-0)]. Si l'appui à la violence politique demeure faible dans l'ensemble, l'événement a néanmoins mis en lumière un clivage générationnel inquiétant : les jeunes Canadiens âgés de 18 à 34 ans sont nettement plus nombreux à considérer que la violence est parfois nécessaire pour provoquer un changement social, une proportion en hausse depuis 2024 [[IU3](#page-45-0)].

### **4. Perceptions of political violence trigger chilling effects on political expression and democratic participation**

Debates surrounding public reactions to Charlie Kirk's death, including condemnation of celebratory comments and high-profile professional sanctions such as the suspension of Jimmy Kimmel, raised important questions about the boundaries of political speech [[IU2](#page-50-0), [IU5](#page-31-0)]. Canadians broadly agree that inciting political violence should carry consequences, including professional sanctions, legal action, or removal from social media, with somewhat lower support for penalizing speech that celebrates or justifies violence. A small majority (56%) are also willing to trade off some civil liberties so that governments can effectively address the threat of political violence [[IU7](#page-17-0)]. At the same time, Canadians generally perceive that freedom of expression is under threat, even though a minority have personally experienced professional, legal, or social consequences for their political speech [[IU7](#page-17-0)]. A high but stable proportion of Canadians do not feel comfortable posting their political opinion online [[IU6](#page-24-0)]. Among those who do, roughly half think about the possibility of harassment or professional, legal, and social repercussions before expressing (or choosing not to express) their views [[IU7](#page-17-0)]. More generally, concern about political violence constrains democratic engagement, with half of Canadians saying they will avoid some political activities, such as putting up political signs, posting opinions online, or attending protests, because of such concern [[IU3](#page-45-0)].

Overall, this incident was classified as minor because, although most Canadians heard about it and it led to misinformation and polarized online debates, including content amplified by Russian state-affiliated actors, these dynamics were relatively short-lived and had limited negative offline consequences.

### **Lessons Learned**

The Charlie Kirk assassination provides critical insights on future instances of political violence that can help us better prevent, mitigate, prepare for, and respond to future events. Our investigation reveals 5 key lessons for consideration.

### **4. Les perceptions de la violence politique et leurs effets dissuasifs sur l'expression politique et la participation démocratique**

Les débats entourant les réactions publiques à la mort de Charlie Kirk, y compris la condamnation de commentaires jugés célébratoires et des sanctions professionnelles très médiatisées, telles que la suspension de Jimmy Kimmel, ont soulevé d'importantes questions quant aux limites de l'expression politique [[IU2](#page-50-0), [IU5](#page-31-0)]. Les Canadiens s'entendent largement pour dire que l'incitation à la violence politique devrait entraîner des conséquences, qu'il s'agisse de sanctions professionnelles, de poursuites judiciaires ou du retrait de contenus des réseaux sociaux, avec un soutien légèrement moindre lorsqu'il s'agit de pénaliser des propos qui célèbrent ou justifient la violence. Une faible majorité (56 %) se dit également prête à accepter certaines restrictions des libertés civiles afin de permettre aux gouvernements de lutter plus efficacement contre la menace que représente la violence politique [[IU7](#page-17-0)]. Parallèlement, les Canadiens estiment généralement que la liberté d'expression est menacée, même si seule une minorité affirme avoir subi personnellement des conséquences professionnelles, juridiques ou sociales en raison de ses opinions politiques [[IU7](#page-17-0)]. Une proportion élevée, mais stable, de Canadiens ne se sentent pas à l'aise de publier leurs opinions politiques en ligne [[IU6](#page-24-0)]. Parmi ceux qui le font, environ la moitié réfléchissent au risque de harcèlement ou de répercussions professionnelles, juridiques ou sociales avant de s'exprimer (ou de choisir de ne pas le faire) [[IU7](#page-17-0)]. Plus largement, les préoccupations liées à la violence politique freinent l'engagement démocratique : la moitié des Canadiens déclarent éviter certaines activités politiques, comme afficher des pancartes politiques, publier des opinions en ligne ou participer à des manifestations, en raison de ces craintes [[IU3](#page-45-0)].

Dans l'ensemble, cet incident a été classé comme mineur, car, bien que la majorité des Canadiens en aient entendu parler et qu'il ait donné lieu à de la désinformation et à des débats en ligne polarisés, notamment des contenus amplifiés par des acteurs affiliés à l'État russe, ces dynamiques ont été relativement éphémères et ont entraîné des conséquences négatives limitées hors ligne.

# **Leçons apprises**

L'assassinat de Charlie Kirk offre une perspective essentielle pour comprendre les futures manifestations de violence politique et améliorer notre capacité à les prévenir, à en atténuer les effets, à nous y préparer et à y répondre. Notre analyse fait ressortir cinq leçons clés à prendre en considération.

### **1. Online sentiments provide a skewed and polarized misrepresentation of offline sentiments.**

As is routinely the case, the online conversation provides a highly distorted view into population-level attitudes. This divide was particularly noticeable during the fallout from the assassination in Canada where online animus, blame, and division was rampant and where politician and news voices were minimal compared to the massive reach of influencers. Meanwhile, the most prevalent offline emotional reactions were sadness/grief and indifference. Both anger and happiness were far more observed online than personally experienced. The infrastructure of online spaces prioritizes hostile, emotive, and contentious framings of events far above reserved and moderate ones.

### **2. Widespread condemnation of political violence does not eliminate its chilling effects on democratic participation.**

The Charlie Kirk assassination prompted near-universal condemnation across platforms and borders, with content creators globally expressing predominantly critical reactions. Yet despite this shared moral consensus, Canadians and Americans alike viewed the assassination as representative of a broader societal problem rather than an isolated event, fears of violence remain elevated, and majorities in both countries expect the problem to worsen. Concerningly, people tend to overestimate the extent to which their political opponents support violence and these dynamics appear to be chilling democratic participation. While consensus against violence is encouraging, the gap between actual attitudes and perceived threats risks eroding the civic engagement that healthy democracies depend on.

### **3. Young people are critical outliers in their online exposure and attitudes toward political violence, warranting focused attention.**

Young Canadians stood out as the demographic most familiar with Charlie Kirk prior to his assassination, and they were significantly more likely than older generations to encounter news of his death through social media rather than traditional broadcast media. This heightened online exposure matters because social media environments tend to amplify polarizing figures and emotionally charged narratives. More concerning, young Canadians are also more likely than older age groups

### **1. Les sentiments exprimés en ligne offrent une représentation biaisée et polarisée des attitudes hors ligne.**

 Comme c'est souvent le cas, les conversations en ligne donnent une image fortement déformée des attitudes à l'échelle de la population. Après l'assassinat, cette fracture a été particulièrement visible au Canada, où l'animosité, les accusations et les divisions dominaient les échanges en ligne, tandis que les voix des responsables politiques et des médias étaient relativement peu présentes comparativement à la portée massive des influenceurs. En revanche, les réactions émotionnelles les plus répandues hors ligne étaient la tristesse, le deuil et l'indifférence. La colère comme la satisfaction ont été beaucoup plus observées en ligne que réellement ressenties à titre personnel. L'architecture même des espaces numériques favorise les cadrages hostiles, émotionnels et conflictuels des événements, au détriment de lectures plus mesurées et modérées.

### **2. La condamnation généralisée de la violence politique n'en neutralise pas les effets dissuasifs sur la participation démocratique.**

L'assassinat de Charlie Kirk a suscité une condamnation quasi unanime, toutes plateformes et frontières confondues, les créateurs de contenu à l'échelle mondiale exprimant majoritairement des réactions critiques. Malgré ce consensus moral partagé, tant les Canadiens que les Américains ont perçu cet assassinat comme le symptôme d'un problème sociétal plus large plutôt que comme un incident isolé. Les craintes liées à la violence demeurent élevées, et des majorités dans les deux pays s'attendent à une aggravation de la situation. Plus préoccupant encore, les individus ont tendance à surestimer le degré de soutien de leurs adversaires politiques à la violence, une dynamique qui semble freiner l'engagement démocratique. Si le rejet largement partagé de la violence est encourageant, l'écart entre les attitudes réelles et les menaces perçues risque néanmoins d'éroder la participation civique dont dépendent les démocraties en bonne santé.

### **3. Les jeunes constituent un groupe distinct, tant par leur exposition en ligne que par leurs attitudes à l'égard de la violence politique, ce qui nécessite une attention particulière.**

Les jeunes Canadiens se distinguent comme le groupe démographique le plus familier avec Charlie Kirk avant son assassinat et étaient nettement plus susceptibles que les générations plus âgées d'apprendre la nouvelle de sa mort par les réseaux sociaux plutôt que par les médias traditionnels. Cette exposition accrue en ligne est significative, car les environnements des réseaux sociaux tendent à amplifier les figures polarisantes et les récits to view political violence as sometimes necessary for social change and this belief appears to be growing over time. While the majority of young people still reject that violence, the combination of greater exposure to polarized online content and rising openness to violence as a political tool represents a warning sign. These patterns underscore the need for interventions that address both the information environments young people inhabit and the validity of democratic channels.

### **4. Major polarizing events follow a brief but intense conversation pattern that includes speculation, blame and misinformation.**

Online discourse following the Kirk assassination surged rapidly across platforms and borders before subsiding within roughly two weeks. The narrative arc followed a recognizable pattern: initial waves of conspiracy theories, grief, and speculative sense-making around the shooter's identity gave way to polarized ideological contestation, with claims about celebrations of Kirk's death generating the sharpest divisions.1 Discussion topics ranged from the shooter's motives and identity to tangential debates about free speech, media responses, and broader cultural grievances. Notably, the most polarizing claims were predominantly refuted rather than supported by users. This pattern mirrors research on mass shootings and crisis events, which documents rapid attention spikes driven by speculation and blame that collapse once initial claims are challenged or the news cycle moves on. For those monitoring information ecosystems, the predictability of these cycles presents both a challenge (misinformation spreads fastest in the early, factpoor window) and an opportunity to prepare evidence-based responses before narratives calcify.

### **5. Russia's efforts to exploit events of political instability continues and exceed that of other countries.**

Russian state-affiliated media and creators demonstrated disproportionate interest in the Kirk assassination compared to other countries with documented foreign influence operations. fortement chargés sur le plan émotionnel. Plus inquiétant encore, les jeunes Canadiens sont également plus enclins que les autres groupes d'âge à considérer que la violence politique est parfois nécessaire pour provoquer un changement social, une opinion qui semble gagner du terrain au fil du temps. Bien que la majorité des jeunes rejettent toujours la violence, la combinaison d'une exposition accrue à des contenus en ligne polarisés et d'une ouverture croissante à la violence comme outil politique constitue un signal d'alerte. Ces constats soulignent la nécessité d'interventions ciblées qui prennent en compte à la fois les environnements informationnels fréquentés par les jeunes et la légitimité des canaux démocratiques.

### **4. Les événements fortement polarisants donnent lieu à des cycles de discussion brefs mais intenses, marqués par la spéculation, l'attribution des responsabilités et la désinformation.**

À la suite de l'assassinat de Charlie Kirk, les échanges en ligne ont rapidement gagné en intensité sur l'ensemble des plateformes et au-delà des frontières nationales, avant de retomber en l'espace d'environ deux semaines. La trajectoire narrative a suivi un schéma bien reconnaissable : des vagues initiales de théories du complot, de deuil et des spéculations notamment autour de l'identité du tireur ont cédé la place à des affrontements idéologiques polarisés, les allégations selon lesquelles la mort de Kirk aurait été célébrée générant les clivages les plus marqués.<sup>1</sup> Les discussions ont porté aussi bien sur les motivations et l'identité du tireur que sur des débats connexes concernant la liberté d'expression, le traitement médiatique de l'événement et des griefs culturels plus larges. Fait notable, les affirmations les plus polarisantes ont majoritairement été réfutées plutôt que soutenues par les utilisateurs. Ce schéma fait écho aux travaux de recherche sur les fusillades de masse et les événements de crise, qui montrent des pics d'attention rapides, alimentés par la spéculation et l'attribution de blâmes, suivis d'un effondrement de l'intérêt une fois les premières affirmations contestées ou lorsque le cycle médiatique se déplace. Pour les acteurs chargés de surveiller les écosystèmes de l'information, la prévisibilité de ces cycles constitue à la fois un défi — la désinformation se propage le plus rapidement dans les premières phases, lorsque les faits sont rares — et une occasion de préparer des réponses fondées sur des données probantes avant que les récits ne se cristallisent.

### **5. Les tentatives de la Russie pour exploiter les épisodes d'instabilité politique persistent et dépassent celles d'autres pays.**

Comparativement à d'autres pays connus pour leurs opérations d'ingérence étrangère, les médias et créateurs affiliés

Russian sources surged well above their already elevated baseline of attention to North American politics, promoting highly politicized and conspiratorial narratives. These tactics align with documented Russian information warfare playbooks: rapid saturation of the information space before facts are established, deployment of conflicting storylines to fragment understanding, and selective amplification of content with the greatest potential to deepen polarization. This pattern is not isolated with similar tactics deployed during moments of Western political instability, from the Trump assassination attempt to high-profile diplomatic confrontations. Recognizing these predictable responses to future crises is essential for building information resilience.

### **The Final Word**

The Charlie Kirk assassination was a minor information incident with outsized visibility, generating intense online polarization that belied the more subdued reactions most Canadians actually experienced. Russian state-affiliated actors exploited the moment predictably, while young Canadians' growing openness to political violence as a tool for change combined with heightened exposure to polarizing voices online warrants particular attention. The incident's legacy is dual: it illuminates democratic vulnerabilities and offers an opportunity to prepare for future active shooter or assassination events, but it also deposits another layer of partisan animosity that accumulates with each such event and proves far harder to dispel than the conspiratorial narratives that accompany it.

à l'État russe ont manifesté un niveau d'attention nettement supérieur à l'égard de l'assassinat de Charlie Kirk. Les sources russes ont porté une attention supérieure à d'habitude à la politique nord-américaine, en diffusant des récits fortement politisés et conspirationnistes. Ces pratiques s'inscrivent dans des stratégies désormais bien documentées de la guerre informationnelle russe : occuper rapidement l'espace informationnel avant l'établissement des faits, multiplier des récits concurrents afin de brouiller la compréhension, et amplifier de manière sélective les contenus les plus susceptibles d'alimenter la polarisation. Ce schéma n'est pas isolé : des tactiques comparables ont été observées lors d'autres épisodes d'instabilité politique en Occident, de la tentative d'assassinat de Donald Trump à des confrontations diplomatiques très médiatisées. Identifier ces réactions prévisibles en amont des crises futures est essentiel pour renforcer la résilience de l'écosystème de l'information.

# **Mot de la fin**

L'assassinat de Charlie Kirk a constitué un incident d'information mineur, mais d'une visibilité disproportionnée. Il a donné lieu à une polarisation en ligne particulièrement marquée, contrastant avec les réactions généralement plus sobres vécues par la majorité des Canadiens. Des acteurs affiliés à l'État russe ont, sans surprise, cherché à tirer parti de la situation, tandis que l'ouverture croissante des jeunes Canadiens à la violence politique comme levier de changement, conjuguée à une forte exposition à des discours polarisants en ligne, mérite une attention soutenue. Le legs de cet incident est double. Il met en évidence certaines vulnérabilités démocratiques et offre une occasion de mieux anticiper et gérer de futurs épisodes de fusillades ou d'assassinats. Il laisse aussi derrière lui une couche supplémentaire d'animosité partisane qui, à l'image d'événements similaires, s'accumule et s'avère beaucoup plus difficile à résorber que les récits conspirationnistes qui l'entourent.

### **Contents**

| Overview<br>9                                                                                                                         |
|---------------------------------------------------------------------------------------------------------------------------------------|
| Incident Update 8<br>Foreign State Creators' Response to<br>Charlie Kirk Assassination<br>11                                          |
| Incident Update 7<br>Canadians' Views on Political Violence and<br>Freedom of Expression: Reality, Risks, and<br>Misperceptions<br>16 |
| Incident Update 6<br>Canadians' Awareness, Beliefs, and<br>Reactions to the Charlie Kirk Shooting<br>23                               |
| Incident Update 5<br>Mapping Controversial and Polarized Claims<br>Following the Assassination of Charlie Kirk 30                     |
| Incident Update 4<br>Shockwaves on TikTok: Understanding the<br>International Reaction to the Assassination<br>of Charlie Kirk<br>39  |
| Incident Update 3<br>Shared Anxiety, Different Lenses: Comparing<br>Canadian and American Perceptions of<br>Political Violence<br>44  |
| Incident Update 2<br>Online Shockwaves: Reactions to the Charlie<br>Kirk Incident on X in Canada and Abroad . 49                      |
| Incident Update 1<br>Kremlin Doppelgänger Tactics and the<br>Exploitation of the Charlie Kirk Murder<br>55                            |
| Notes<br>59                                                                                                                           |

### <span id="page-10-0"></span>**Overview**

# **Ripple effects of the Charlie Kirk Assassination in the Canadian Information Ecosystem**

**What happened:** On September 10, 2025 prominent American right-wing activist and founder of Turning Point USA, Charlie Kirk, was assassinated while at a public rally on the campus of Utah Valley University.

**Impact:** His death has resulted in a surge of highly charged and polarized online conversations, disinformation and calls for political violence in the American but also Canadian information environment. Canadian [commentators have been targeted](https://www.thetyee.ca/Opinion/2025/09/15/Dangerous-Targeting-Canadian-Journalist/) and Canadian politicians from across the political spectrum have weighed in, warning about the dangers of political violence and ideological disagreement. Foreign states, such as [Russia,](https://apnews.com/article/charlie-kirk-russia-china-disinformation-putin-trump-bce0174644351c70811ae4a847ffa767)  [China, and Iran](https://apnews.com/article/charlie-kirk-russia-china-disinformation-putin-trump-bce0174644351c70811ae4a847ffa767), have demonstrated interest in amplifying and intensifying [the polarizing discourse](https://x.com/mfa_russia/status/1966495830241608123?s=46&t=DoWUMlICe1rAToovv2mgQA).

**Why it matters:** The assassination has been followed by a massive amount of polarized online discussions in the Canadian information ecosystem, fueling disinformation, partisan conflict, and mobilization.

# **What happened?**

**Incident Classifications**: Disinformation, misinformation, domestic manipulation, USA, foreign influence, polarization, extremism and/or radicalization online, conspiracy theories

**Origin**: The incident originated on September 10, 2025, in the United States, when prominent right-wing activist Charlie Kirk was shot and killed during a public event at Utah Valley University. At the moment of the shooting, Kirk was engaged in a debate about mass shootings with a member of the public, characteristic of his career as founder of Turning Point USA, where he would venture across the USA to university campuses and engage in political debates on contentious issues including abortion, women's rights, race, LGBTQA+ rights, Ukraine-Russia, and Israel-Palestine. These interactions were usually filmed and

uploaded to his social media accounts, where online discourse would often center on particularly sharp and sensational small clips of the conversations. Following this incident, politicians and international figures condemned the act of political violence. While supporters of Kirk posted tributes of Kirk and framed him as a martyr and advocate for free speech, other critics pointed to his previous remarks to challenge his commemoration. This has led to increasing domestic and international polarization, calls for rebellion and violence, disinformation, and has potentially fueled radicalization online. Individuals who have spoken out about the matter have been targets (some have been [investigated by their employers and lost their jobs](https://www.cbsnews.com/news/charlie-kirk-shooting-employees-fired/), for example), leaving a chilling effect on free speech.

These tensions have spread into Canada, increasing growing concern over free speech, culture wars, and political violence. A University of Toronto professor [has been placed on leave](https://ici.radio-canada.ca/rci/en/news/2192494/u-of-t-professor-placed-on-leave-after-commenting-on-charlie-kirks-shooting-death) after comments related to Kirk's shooting, [an Ottawa man was charged](https://www.cbc.ca/news/canada/ottawa/ottawa-man-charged-after-school-threatened-over-charlie-kirk-post-1.7633754)  after making posts about Kirk and threatening a school, and vigils across Canada for Kirk (e.g. Winnipeg, Calgary, and Regina) [have been, in some cases, met with protestors.](https://www.mapleridgenews.com/news/vigil-for-charlie-kirk-in-maple-ridge-also-draws-protesters-8253035)

**Motive and Tactic:** Though the motive of the attack remains under investigation, Canadian and international commentators have interpreted the event as a [political assassination](https://www.reuters.com/world/us/prosecutors-seek-death-penalty-charlie-kirks-accused-assassin-2025-09-16/). Prominent Canadian politicians including Danielle Smith and Pierre Poilievre have stressed that Kirk was killed for his views, echoing these claims of ideological persecution.

**Scope and Reach**: One week after the event, the story continues to be prominent in the Canadian online conversation. This incident has also garnered international media attention and public interest. Beyond Canada, individuals in other countries including the [UK](https://www.independent.co.uk/news/uk/home-news/charlie-kirk-shooting-vigil-london-b2824533.html?utm_source=chatgpt.com), [Spain](https://americanfaith.com/madrid-vigil-charlie-kirk-assassination/?utm_source=chatgpt.com), and [Australia](https://www.dailytelegraph.com.au/subscribe/news/1/?sourceCode=DTWEB_WRE170_a_GPT&dest=https%3A%2F%2Fwww.dailytelegraph.com.au%2Fnews%2Fnsw%2Fcrowd-gathers-to-mourn-slain-conservative-charlie-kirk-in-sydney%2Fnews-story%2F08fb6aeb5d54811d842650b12d21a5bf&memtype=anonymous&mode=premium&v21=GROUPA-Segment-2-NOSCORE) have gathered for vigils in remembrance of Kirk. Online Canadian news sites including [Juno news](https://www.junonews.com/p/candice-malcolm-the-legacy-media) and [Rebel News](https://www.rebelnews.com/saskatoon_subreddit_shuts_down_post_inquiring_about_charlie_kirk_vigil) have covered the criticisms of progressive Canadians' responses and pointed to potential hostility towards conservatives. Previous videos of Kirk commenting on Canadian politics [have re-surfaced](https://torontosun.com/news/national/charlie-kirks-blunt-assessment-of-life-in-canada-goes-viral-after-shooting-death). Prominent foreign state media organizations, including Russia, China and Iran, have also [seized on the political moment](https://www.newsguardrealitycheck.com/p/russian-chinese-and-iranian-state) to push polarizing narratives and divisive international politics and [generally exploit the events](https://www.washingtonpost.com/technology/2025/09/18/russia-leverages-kirk-shooting-propaganda-false-claims-surge/).

**Impact:** This incident intensifies concern over whether political division and violence may increase [in Canada.](https://www.thestar.com/opinion/contributors/charlie-kirks-assassination-is-a-warning-shot-for-canada-heres-why/article_e2056a0a-de7d-45bf-9e37-c809c5542e9d.html) Many voices point to threats against free speech and ideological intolerance while also cautioning against increasing radicalization. The perpetrators' potential ties to online cultures (e.g. memetic inscriptions on bullet casings) illustrate concern for online platforms as a source of radicalization.

Click [here for a detailed timeline of events](https://docs.google.com/document/d/1yDf4Hkp5UcZDhoPwJfplULpktqXzkfwSgGOsEveO1CE/edit?tab=t.0).

### **Why does this matter?**

The assassination of Kirk underscores the threat of potential spillover of political violence into Canada. Canada's media ecosystem and political landscape is intermittently tied to the United States, often reactionary to the political affairs south of the border. Acts of violence in the US can influence domestic polarization and radicalization and threats of violence and uncertainty. Polarization has increased online, with Canadian outlets such as True North Wire and Rebel News critiquing progressive Canadians' responses to Kirk's death. Meanwhile, progressive and left-leaning commentators point to the controversy around Kirk as a political figure and the policies he had advocated for. Commentators such as Rachel Gilmore have been at the [centre of online harassment and threats](https://thetyee.ca/Opinion/2025/09/15/Dangerous-Targeting-Canadian-Journalist/) for her comments about Kirk. A Canadian Conservative MP Andrew Scheer has even chimed in, claiming Gilmore as 'twisted' and had 'so much hate in her'. Canadian politicians, including Pierre Poilievre and Danielle Smith have framed the death of Kirk as a warning against ideological prosecution, tying the incident to partisan differences. This illustrates how US polarization is pulled into the forefront of Canadian political discourse. Further, increasing online partisan polarization and efforts to dox individuals who speak critically about Kirk reveal the tenuous nature of free speech and a potential chilling effect in the Canadian media ecosystem.

This event has also sparked increased false information, disinformation, and conspiracy theories that enter the Canadian media ecosystem. This includes misleading narratives about the shooter's motives and background and the security of the Utah Valley University event. For example, [information regarding the](https://news.sky.com/story/charlie-kirk-suspect-tyler-robinson-was-in-romantic-relationship-with-transgender-roommate-utah-governor-13431020)  [shooter's connections to the trans community](https://news.sky.com/story/charlie-kirk-suspect-tyler-robinson-was-in-romantic-relationship-with-transgender-roommate-utah-governor-13431020) and [the nature of](https://www.theguardian.com/us-news/2025/sep/16/charlie-kirk-shooting-prosecutor-utah)  [the relationship with his roommate](https://www.theguardian.com/us-news/2025/sep/16/charlie-kirk-shooting-prosecutor-utah) can be used to reinforce existing and harmful stereotypes that trans individuals are violent. Inaccurate information and disinformation can be used as a political tool in the Canadian context to shape public perception and opinion related to public safety and social issues by exploiting grievances.

Recent vigils held across Canada and attended by thousands of Canadians demonstrate the deep ties between US and Canadian political culture. It illustrates how Kirk has significant influence amongst right-wing politics in Canada, underscoring the potential for transnational right-wing mobilization. Previous footage of Kirk's comments about Canada have also resurfaced and received increased engagement. This demonstrates the influence of the online ecosystem in amplifying right-wing politics across the border and mobilizing Canadians to gather in collective action.

### **What are we doing about it?**

This notification launches an investigation, led by the [Canadian](https://www.cdmrn.ca/)  [Digital Media Research Network \(CDMRN\)](https://www.cdmrn.ca/), into the online responses to the assassination of Charlie Kirk and overall impacts to the Canadian media ecosystem.

### **Key questions we are focusing on:**

**Content:** What are the major claims, narratives, and mobilizations that have emerged in Canadian online conversations in the aftermath of Charlie Kirk's assassination? How are different groups framing the events?

**Network:** Which Canadian influencers, media outlets, and political figures have focused on Kirk's assassination? How are their claims and narratives being shared, supported and contested across different online communities? To what extent is the Canadian conversation being driven by international actors?

**Impact:** What level of engagement have these claims and narratives received in the Canadian information ecosystem? Have they meaningfully shaped political discourse in Canada on issues such as free speech, extremism, culture wars, and polarization?

The CDMRN and research partners will address these questions through a combination of techniques including survey and social media analysis. In the coming weeks, we will release incident updates on these and, where appropriate, other topics.

### <span id="page-12-0"></span>**Incident Update 8**

# **Foreign State Creators' Response to Charlie Kirk Assassination**

**Aengus Bridgman Media Ecosystem Observatory**

### **Key takeaways**

- 1. **Russia devoted far more attention to the event than China or India, relative to their typical coverage of North American politics:** Russian state-affiliated creators increased their already high attention to North American topics by about 13 percentage points after the assassination, indicating a strategic focus on this event, whereas Chinese and Indian state creators' coverage remained consistent with their usual attention to North America.
- 2. **Russia stood out for advancing overtly polarized narratives:** Russian state-affiliated creators promoted **highly politicized and conspiratorial claims and interpretations of the events**, while Indian state-affiliated creators focused on factual reporting without editorializing, and Chinese state-affiliated creators focused on democratic (in)stability and violence in the U.S.
- 3. **Russia's response to the Kirk assassination fits a broader, recurring pattern in its coverage of major U.S. events**: Across multiple moments, including the Trump assassination attempt, the 2025 U.S. election, the LA wildfires, and the first Trump-Zelensky meeting in the White House, state-affiliated creators consistently used U.S. news events to advance narratives of Western dysfunction, hypocrisy, and decline. The response to the Kirk assassination is representative of a stable communicative strategy in which Russia exploits crises to amplify conspiracy theories, polarization, and institutional distrust.

### **Introduction**

### **Context**

Recent years have seen growing recognition that [foreign states](https://www.cdmrn.ca/publications/kremlin-doppelgnger-tactics-and-the-exploitation-of-the-charlie-kirk-murder)  [exploit social](https://www.cdmrn.ca/publications/kremlin-doppelgnger-tactics-and-the-exploitation-of-the-charlie-kirk-murder) and political crises in Western democracies to advance influence operations and exacerbate domestic polarization. For example, the National Security and Intelligence Committee of Parliamentarians (NSICOP) in Canada devoted much of its 2024 "[Special Report on Foreign Interference in Canada's](https://nsicop-cpsnr.ca/reports/rp-2024-06-03/special-report-foreign-interference.pdf)  [Democratic Processes and Institutions](https://nsicop-cpsnr.ca/reports/rp-2024-06-03/special-report-foreign-interference.pdf)" to describing how state actors use covert and deceptive tactics including disinformation, online manipulation, and interference targeting diaspora communities to undermine trust in democratic institutions and public discourse. A high-profile event such as the assassination of Charlie Kirk may have provided a key opportunity for state-affiliated outlets to produce and spread disinformation campaigns. Early reports have widely reported that to be the case in this instance including [Politico](https://www.politico.com/news/2025/09/17/charlie-kirk-shooting-disinformation-russia-china-iran-00570357), [Newsweek](https://www.newsweek.com/charlie-kirks-assassination-governor-cox-bots-adversaries-violence-2128699) and [ABCNews](https://abcnews.go.com/Politics/security-analysts-flag-rise-russian-created-misinformation-posts/story?id=125640078). This update uses social postings by known state-affiliated creators (media, organizations, and influencers) to evaluate the extent to which Russia, China, and India (three major foreign influence actors) contributed to the spread of disinformation.

### **Key questions**

This report evaluates the following research questions:

- 1. What was the volume, timing, and engagement level of select foreign state-affiliated creators' coverage of Charlie Kirk's assassination?
- 2. How did state-affiliated creators from China, Russia, and India frame Charlie Kirk's assassination?
- 3. Was this pattern of attention and framing unusual?

# **Approach and Considerations**

This analysis examines social media posts from 116 state-affiliated entities associated with China, Russia, and India. Accounts were identified based on platform disclosures2 (where platforms have identified country of origin), self-identification as state-affiliated entities, and/or verification of their affiliation using independent media experts who monitor these countries' influence operations. The dataset comprises 939,883 posts collected across multiple platforms including X (formerly Twitter), Facebook, Instagram, YouTube, Telegram, and TikTok, covering the period from January 1, 2023 through September 23, 2025.

By "state-affiliated creators," we refer to organizations and individuals that receive funding from, are editorially influenced by,

<span id="page-13-0"></span>**Figure 1. State-affiliated media coverage of the United States and Canada: September 2025 vs. baseline**

or operate under the direct authority of national governments. This includes entities explicitly labeled as state media (such as RT, CGTN, and DD News) as well as organizations and individuals with documented state connections through ownership structures, editorial oversight, or financial dependencies. We acknowledge that the degree of state affiliation varies across entities, ranging from direct government operation to more indirect forms of influence.

We employ topic modelling to assess the posts of these creators and evaluate the extent to which these focused on the United States, Canada, and what sort of frames they employ to discuss the Charlie Kirk assassination. See [methodology](#page-16-0) for additional information about the approach employed.

### **1. What was the volume, timing, and engagement level of foreign state-affiliated creators' coverage of Charlie Kirk's assassination?**

Russian, Indian, and Chinese creators have significant and continuous coverage of the United States. As shown in the top plot of [Figure 1](#page-13-0), the three-year baseline (2023–2025) showing attention among top topics to the United States (and Canada) ranges from 18% (India) to 29% (Russia). Canadian coverage is far lower than that of the United States and is comparably low across all three countries, although India does tend to cover Canadian stories at a slightly higher frequency (1.5% for India) than the others (1.2% for China and Russia). This high baseline attention to events in the United States is critical to understanding how state media might respond to potentially polarizing events.

The assassination of Charlie Kirk on September 10, 2025 produced a dramatic and immediate shift in attention toward the United States. In the days preceding the assassination, all three countries were tracking below their historical baselines, with Russian topic coverage at around 18%, well below its typical 29% level. Over the next several days, Russian state media topic coverage of North America soared to over 40%, representing a more than doubling of attention and substantially exceeding baseline levels. This spike was sustained for approximately one week before beginning to decline. Indian and Chinese state media showed more modest responses. Both increased their North American coverage following September 10, with India rising from 13% in the 2-weeks before to 28% and China from 15% to 28%. Notably, these post-assassination levels brought both countries roughly in line with or just above their historical

**Figure 2. Examples of conspiratorial or polarizing claims pushed by Russian state-affiliated creators**

baselines, suggesting the Kirk assassination generated sufficient news interest to temporarily elevate coverage to typical levels rather than producing exceptional attention.

The differential response across countries is revealing. Russia's exceeded its already high baseline by roughly 13 percentage points suggests either heightened editorial interest in US political violence or strategic prioritization of this particular event. In contrast, China and India's responses appear more consistent with routine news processing: coverage increased to meet baseline expectations but did not dramatically exceed them. The same patterns hold when looking at engagement rather than post volume.

### **2. How did state-affiliated creators from China, Russia, and India frame Charlie Kirk's assassination?**

To better understand the response, we manually reviewed the 100 most engaged with posts about the assassination and fallout from each of China, Russia, and India to assess the extent to which the coverage was factual and news-oriented, or focused on conspiracies and polarizing claims.

Russian sources, particularly RT, adopted the most politically charged framing, systematically amplifying instances of Western figures celebrating Kirk's death from a [New York comedian](https://x.com/i/status/1966131191620456863)  [mocking the murder](https://x.com/i/status/1966131191620456863) to a British rapper's [profane tribute](https://x.com/i/status/1967168382337450098). RT's highest-engagement post (4,127 interactions) featured international solidarity ("[We are Charlie Kirk](https://x.com/i/status/1966911371959816573)" from South Korea), while the Russian Foreign Ministry's official statement explicitly framed the killing as "[yet another manifestation of the entrenched](https://x.com/i/status/1966532261630628252) 

[Western ultraliberal view — one postulating that violence and](https://x.com/i/status/1966532261630628252)  [terror are acceptable means of dealing with opponents](https://x.com/i/status/1966532261630628252)." Russian coverage also promoted conspiratorial questioning of mainstream coverage of the event, with one viral post demonstrating that the rifle allegedly used ["does not even fit"](https://x.com/i/status/1968619260466319429) in a backpack as claimed, asking viewers "Do you buy the official Tyler Robinson story?" Several prominent Telegram posts framed Kirk's death apocalyptically as a "Turning Point" killed by "[liberals,](https://t.me/duaboroda/23259)  [globalists, trans people and Soros agents](https://t.me/duaboroda/23259)."

Indian state-affiliated media took a different approach, functioning primarily as a high-volume news aggregator. India's top posts were dominated by straightforward video news coverage from outlets like FirstPost, India Today, Times of India, and Republic World, with the highest-engagement content simply showing the [moment of the assassination](https://www.instagram.com/p/C_KTX8mPq4R). Indian coverage emphasized factual elements including the [FBI investigation](https://www.instagram.com/p/C_RSxGYP7Mi), suspect identification, and [memorial services](https://www.instagram.com/p/C_Yw2rrP6vN), without the political commentary prevalent in Russian content.

Chinese state media produced the least coverage, with content from CGTN and South China Morning Post maintaining a neutral, wire-service tone. China's coverage focused on factual reporting of events (e.g., [Trump ally Charlie Kirk shot dead at Utah](https://www.youtube.com/watch?v=DgDph-3me_Q)  [event](https://www.youtube.com/watch?v=DgDph-3me_Q)) and when they did explore the broader context of American political violence: "[What's behind the surge in violence in U.S.](https://www.youtube.com/shorts/rJy0yKzoC-s)  [politics?](https://www.youtube.com/shorts/rJy0yKzoC-s)", it was framed as an American domestic issue. Chinese sources, when editorializing, emphasized concerns about [U.S.](https://www.youtube.com/watch?v=nloL0SLAUEY)  [political polarization](https://www.youtube.com/watch?v=nloL0SLAUEY) and free speech implications (particularly around the Jimmy Kimmel suspension), positioning itself as an observer of American dysfunction.

**Figure 3. State-affiliated coverage of the United States and Canada: Six important political events**

### **Was this pattern of attention and framing unusual?**

To determine whether the heightened attention and polarizing framing from Russia to this event reflects a broader communicative pattern, we performed the same time analysis of five other significant events: 1) the assassination attempt of Trump on July 13, 2024; 2) the election of Donald Trump on November 5, 2025; 3) the Los Angeles Wildfires in January 2025; 4) the election of Mark Carney in Canada on April 28, 2025; 5) the Trump-Zelensky meeting in the White House on February 14, 2025.

The results demonstrate a consistent pattern where major political events in the United States prompt increased attention from Russia, India, and China. Note that the only Canadian event, the Canadian Federal Election, had some engagement several days earlier following the Lapu-Lapu Day festival where 11 people were killed in Vancouver. Those events actually received more attention than Carney and the Liberal Party of Canada winning the election, before Russian attention actually pivoted to anticipation/preparation for Victory Day celebrations. Put this way, the uptick in coverage of the Kirk assassination was relatively modest compared to other major events.

We did a similar review of Russian coverage and particularly the polarization and conspiratorial ideation across the six western-oriented events. For the Trump assassination, Russian coverage emphasized conspiracy theories about "[deep state](https://youtube.com/watch?v=oLQg7xzQGJM)" involvement and CNN's [initial downplaying](https://x.com/BowesChay/status/1812257100992487674) of the event. For the Trump election, coverage adopted a triumphalist tone, highlighting liberal "[meltdowns](https://x.com/blinova14/status/1854015437106782325)" and European [establishment panic](https://x.com/MedvedevRussiaE/status/1855550409571508621). For the LA Wildfires, Russian creators focused on governance failure narratives, with RT and affiliated accounts emphasizing empty [fire hydrants](https://x.com/BowesChay/status/1877301479129731551), the California governor's ([Gavin Newsom](https://x.com/RT_com/status/1878705387416990098)) water mismanagement, [DEI policies blamed for incompetence](https://x.com/RT_com/status/1878913381853184108), and [third-world comparisons](https://x.com/BowesChay/status/1876996147551555595) to LA's fire and water infrastructure. For the only Canadian event, the Lapu-Lapu Day festival saw [neutral](https://x.com/RT_com/status/1916458303955112252), [respectful](https://x.com/RussianEmbassyC/status/1916548234388111368), and [fact-oriented coverage](https://x.com/RT_com/status/1916642834591154659) which suggests that, while [Canada also is subject to the same phenomena](https://www.cdmrn.ca/publications/the-canadianinformation-ecosystem-during-the-2025-federal-election), Russian state-affiliated creators generally focus much more heavily on the United States.

Across all events, Russian state media employed a consistent underlying thesis: American/Western dysfunction, hypocrisy, and moral decay. There is a consistent pattern of Russian response to major U.S. news events that provide opportunities to push that narrative.

### **Conclusion**

The assassination of Charlie Kirk prompted a rapid but uneven response across foreign state-affiliated media ecosystems, revealing important distinctions in how Russia, China, and India use moments of U.S. political crisis. Russia exhibited the most pronounced and concerted shift, dramatically increasing its attention to the United States and deploying overtly politicized, conspiratorial, and emotionally charged narratives that amplified polarization and questioned institutional legitimacy. India, by contrast, largely adopted a high-volume but straightforward news-reporting posture, emphasizing factual updates and avoiding explicit editorialization, while China maintained minimal, restrained coverage focused on political polarization and broader implications for democratic stability.

Taken together, these findings demonstrate that while major U.S. political events reliably attract attention from foreign state-affiliated creators, only Russia consistently uses such moments to amplify divisive and conspiratorial narratives. This distinction has meaningful implications for understanding how different states instrumentalize Western crises as opportunities to escalate polarized information environments.

# <span id="page-16-0"></span>**Methodology**

To identify recurring themes and topics within this corpus, we employed BERTopic, an unsupervised topic modeling technique that leverages transformer-based sentence embeddings to cluster semantically similar documents. Specifically, we used pre-computed embeddings from the paraphrase-multilingual-MiniLM-L12-v2 model, which supports multilingual content and captures semantic meaning across linguistic contexts. Put simply, this method groups documents that are similar in meaning, even if they use different words or are written in different languages. The top-30 topics were identified for each week and each source country (China, Russia, India), ranked by the number of posts classified within each topic.

To generate human-readable topic descriptions, key claims, country associations, and sentiment classifications, we employed the Qwen 3:8b large language model. For each topic, the model analyzed representative posts and generated: (1) a concise title summarizing the topic's main theme, (2) a brief description elaborating on the topic's content, (3) a list of countries mentioned or relevant to the topic, (4) key claims or narratives present within the topic, and (5) an overall sentiment classification (positive, negative, or neutral).

Country associations were extracted through this automated process, with topics classified as concerning the United States or Canada if either country appeared in the generated country list. All topics for the month of September were validated by a human coder and none were changed.

Note that the topic modeling process involves simplifying complex document information grouping similar documents that introduce some subjectivity in topic boundaries. Different modeling parameters or time windows might yield alternative topic structures. Second, our country detection relies on language model inference, which may occasionally misclassify topics or overlook implicit country references. Third, the analysis focuses primarily on content production rather than reach or impact. Note that the accounts included in this analysis represent a considerable but incomplete subset of state-affiliated creators and their presence on social media platforms. Finally, while we observe patterns in topic prevalence and country mentions, causal claims about coordination, strategic intent, or effects on public opinion are beyond the scope here. These findings should be interpreted as descriptive patterns in state-affiliated creators' production rather than evidence of coordinated influence operations.

### <span id="page-17-0"></span>**Incident Update 7**

# **Canadians' Views on Political Violence and Freedom of Expression: Reality, Risks, and Misperceptions**

**Eric Merkley Media Ecosystem Observatory & University of Toronto**

### **Key takeaways:**

- E **Canadians' support for political violence is extremely low, though not for all types of political violence.** Between 3–4% of Canadians report support for egregious acts of political violence by those who share their political views, like arson and assault with a deadly weapon. Support rises slightly for assault (9%) and vandalism (12%) and is considerably higher for harassment (25%).
- E **Canadians tend to exaggerate support for political violence among their political opponents, and these exaggerated perceptions are themselves associated with support for political violence.** On average, Canadians accurately perceive levels of support for political violence among different partisan and ideological groups in society, but tend to overestimate support for political violence by 100% or more among their political opponents. People with the most exaggerated beliefs report between 2–3 times higher support for political violence.
- E **Most Canadians believe there should be consequences for advocating for political violence but are divided on how severe those consequences should be.** 56% of Canadians are willing to tradeoff civil liberties so that governments can address the threat of political violence. About one-third of Canadians believe there should be professional sanctions or legal prosecution for those who justify or celebrate political violence, which rises to 43–53% for those who incite political violence. Still, 20% of Canadians believe there should be no consequences for those who incite political violence, rising to 24–28% for those who celebrate or justify such violence.

- E **Most Canadians perceive freedom of expression to be under threat but are divided on how often they consider the consequences of speech when they post opinions online.** 56% of Canadians believe freedom of expression is under serious or moderate threat in Canada. Among those who post about politics online, roughly half think about possible harassment or the professional, legal, and social consequences of doing so before expressing (or failing to express) their opinions online.
- E **An overwhelming majority of Canadians have not faced legal, professional, or social consequences for expressing their political opinions.** 79% of Canadians report suffering no consequences for expressing political opinions. Only 3–4% have faced legal or professional consequences, while a slightly higher proportion report social conflict or harassment resulting from their speech (8–10%). These patterns may not indicate the absence of risk, but could also reflect chilled speech, with some avoiding political expression due to concern about possible repercussions.

### **Introduction**

### **Context**

Polarization is rising in both Canada and the United States. People are becoming increasingly resentful of people they disagree with politically, a phenomenon known as *affective polarization*. At the same time, acts of political violence have been on the rise in the United States, leading to concerns this trend could spill over to Canada. In [IU3](https://www.cdmrn.ca/publications/comparing-canadian-and-american-perceptions-of-political-violence), we found that while Canadians perceive the threat as less severe than in the United States, half still consider it a serious issue at home, and a third expect it to worsen in the upcoming years.

Combating political violence and the incitement of such violence has implications for civil liberties and freedom of expression – itself a growing area of contention as polarization fuels a trend towards cancel culture, where people impose professional, legal, or social consequences for speech that is deemed unacceptable. Studying how Canadians make tradeoffs between opposing and combating political violence and protecting freedom of expression can inform approaches to prevent the spread of political violence north of the border.

### **Key questions**

This report evaluates the following questions:

1. How high is support for political violence in Canada, and how does it vary depending on the severity of the crime?

<span id="page-18-0"></span>**Figure 1. Support for political violence on 0–10 scale. 0 means not at all justified, 5 means moderately justified, 10 means extremely justified.**

- 2. Do Canadians overestimate their political opponents' support for political violence? Are these misperceptions associated with their own willingness to endorse such actions?
- 3. What consequences do Canadians support for those who incite or justify political violence?
- 4. Do Canadians perceive freedom of expression to be under threat in Canada? Have they felt this threat in their own lives?

### **Approach & Considerations**

The data presented in this update come from two nationally representative surveys: a first of 1,496 Canadians fielded between September 30th and October 4th, 2025, and a second of 1,503 Canadians fielded between October 30 and November 4, 2025.

The second survey contains a pair of experiments. The first experiment, used to evaluate support for political violence, exposes respondents to one randomized scenario about an individual sharing their political views targeting political opponents with either harassment, vandalism, assault, arson, or assault with a deadly weapon (in the form of a car attack). Each respondent only received one scenario (the exact wording is provided in the Methodology section at the end). Respondents were asked about their level of support for the attack in their randomized scenario and whether the perpetrator should serve jail time. For each of our questions

asking about support for political violence, we restrict our sample to the 78% of respondents who were able to recall the nature of the violence in the vignette because prior research has found that support for political violence is inflated by respondent inattention.

The second experiment asks respondents "Which of the following actions, if any, do you believe are justified in response to public comments made about a violent political incident when those comments…" where the nature of the comments was randomized. Respondents either answered this question about speech that "justifies why the event occurred", "celebrates or expresses approval of the event" or "encourages or promotes further violence."

When describing the results, we will often draw attention to the "median" value among our survey respondents. This represents the response of the respondent in the "middle" of the data set. It is a useful value to track for questions when extreme responses may distort the calculation of average or mean responses.

### **1 How high is support for political violence in Canada, and how does it vary depending on the severity of the crime?**

As shown in [Figure 1](#page-18-0), support for political violence is extremely low in Canada. When asked to place how justified it is "for people who share my political values and beliefs to use violence in advancing political goals these days" on a 0–10 scale, as illustrated in [Figure 1](#page-18-0), the respondents reported an average value of 0.8 with a median response of zero where 0 means

**Share of respondents supporting violent action (left) and opposing jail time for the perpetrator (right), with 95% confidence intervals.**

"not at all justified" and 10 means "extremely justified." 84% of respondents placed themselves at either 0 or 1 on that scale and 5% placed themselves above 5. In separate survey questions (not shown in [Figure 1](#page-18-0)), 8% of respondents reported it was at least *occasionally* "OK for people who share my political values and beliefs to send threatening and intimidating messages to political opponents", while 9% of respondents reported it was "OK for people who share my political values and beliefs to harass political opponents on the internet in a way that makes them feel frightened."

These low numbers contrast with our [IU3](#page-45-0) finding that 21% of Canadians believe that political violence is sometimes necessary for social change, a view that reflects a broader, more distant judgment about the world and society. When specifically asked to evaluate whether people who share their political values are justified in using violence or intimidation, support drops dramatically. In other words, some Canadians may see political violence as sometimes necessary in the abstract or when it's part of distant conflicts (e.g., people challenging repressive conditions abroad), but very few are willing to endorse violent or threatening actions when these are tied directly to their own political in-group.

Respondents were also asked about their level of support for a violent politically-motivated action and whether they believed the perpetrator deserved jail time. To understand how perceptions change based on the severity of the crime, each respondent was randomly shown a scenario involving one of five acts of violence: harassment, vandalism, assault with rocks, arson, or an attack with a deadly weapon. Figure 2 shows that support was low for the most serious acts: 3.4% indicated some support for the car attack and 4.3% supported arson, mirroring levels of support found in [other democracies](https://americaspoliticalpulse.com/citizens/international/#citizen-international-intro). Support is slightly higher for assault with rocks (9%) and vandalism (12%), and highest for harassment (25%).

In line with these low levels of approval, only a minority of Canadians oppose jail time for each of these crimes: 1–2% of respondents oppose jail time for the perpetrator for committing arson and a car attack, compared to 17% for assault with rocks, 30% for vandalism, and 26% for harassment.

### **2 Do Canadians overestimate their political opponents' support for political violence? Are these misperceptions associated with their own willingness to endorse such actions?**

While support for political violence is overarchingly low, there is variation in perceptions of support between partisan groupings. When asked where they think the average supporter of different parties and ideologies would place themselves on 0–10 scales of support for political violence ([Figure 3](#page-20-0)), respondents placed far left-supporters at 1.6, the NDP and Liberals at 1, the Conservatives at 1.4, and far-right supporters at 2.2. Trump supporters were perceived to be the most accepting of violence at 2.5. The median response in all cases was zero (the median response is the middle value in the data set).

While perceived support is low, there is some evidence that perceptions of political support are exaggerated for political

<span id="page-20-0"></span>**Figure 3. Perceived support for political violence on 0–10 scales across partisan and ideological groups.**

<span id="page-20-1"></span>**Figure 4. Effect of perceived out-party support for political violence on percent of people supporting violent action in vignette (left) and average support for political violence on 0–10 scale (right). 95% confidence intervals.**

opponents—specifically left leaning voters overestimates the extent right-leaning voters support political violence more than the right does the left. As shown in [Figure 3](#page-20-0), Liberal and NDP supporters place Conservatives at 1.8 and 2.2 on the 0–10 scale, respectively, while Conservatives place themselves at 1.0, roughly 2× the true levels of support shown by Conservatives. Conservatives place Liberal supporters at 1.5 on the scale, while Liberals place themselves at 1, a 50% difference. Liberal and NDP supporters place far-right supporters at around 3 on the 0–10 scale, while Conservatives place them at 1.4. Placement of farleft supporters does not vary much by partisanship (between 1.6 and 1.9 on the 0–10 scale).

Perceiving political opponents as supportive of violence is strongly correlated to one's own support for violence. As shown in the left panel of [Figure 4](#page-20-1), an individual who believes outparty supporters place themselves at 10 (extremely justified) on the support for political violence scale supports the violence contained in the vignette at more than twice the rate of those who place their opponents at 1 on this same scale (20% vs. 9.7%). Similarly, as shown in the right panel of [Figure 4](#page-20-1), a person who places their political opponents at 10 on the scale, places themselves at 2.3 on the same scale. In contrast, those who believe out-party supporters place themselves at 1 on this scale, place themselves at 0.8 on this same scale.

**Figure 5. Percent of respondents supporting four types of consequences for people who either justify, celebrate, or encourage political violence** 

### **3. What consequences do Canadians support for those who incite or justify political violence?**

Canadians tend to report some support for combating political violence. 59% of respondents reported that "the government should take all necessary steps to prevent political violence even if it means infringing on some civil liberties", while 41% agreed with the position that the "government should not infringe on civil liberties, even if it means accepting a greater risk of political violence. These findings suggest that Canadians give slight priority to preventing political violence over safeguarding civil liberties, highlighting the complex trade-offs policymakers have to consider.

To better understand Canadians' opinions on how society should punish political violence, we asked respondents to indicate what consequences would be appropriate for people who either justify, celebrate, or encourage political violence (which was randomly assigned to respondents). Respondents could select any number of consequences from a list that included, among others, professional sanctions (firing, suspension), legal prosecution, inclusion on a government watch list, a ban from social media platforms, and "none".

Across all consequence types, support for sanctions is higher when speech encourages violence than when it celebrates or justifies it. Professional sanctions are supported by 30–34% of Canadians for those who justify or celebrate violence, compared with 43% for those who encourage violence. Similarly, between 32 and 34% of people believe legal prosecution is appropriate for those who justify or celebrate violence, which rises to 53% for those who encourage violence. A sizeable proportion of Canadians believe there should be no consequences for such speech – 24–28% when the speech celebrates or justifies violence, dropping to 20% for the incitement of violence.

The tolerance for professional consequences for speech justifying or celebrating violence somewhat depends on the position of the person in question. As shown in Figure 6, 58% of Canadians believe an elected politician should lose their job, which drops to 46–47% for journalists, public employees, and professors. Only 33% believe a private employee should lose their job in this situation. These numbers are slightly higher than those observed by the [Polarization Research Lab](https://www.dropbox.com/scl/fi/crd3q2gdz06izqoew7vqs/political_violence_toplines.pdf?rlkey=s6sf82xq77eh7mo2ofhhl98xh&e=5&dl=0) in the United States, where support for firing was 36–38% for journalists, public employees, and professors, and 31% for private employees.

### **4. Do Canadians perceive freedom of expression to be under threat? Have they felt this threat in their own lives?**

When asked about the degree to which they believe freedom of speech is threatened in Canada today, 56% say it is seriously or moderately under threat, though only 19% believe it is seriously under threat.

To examine how individuals assess the personal costs of posting

Figure 6.
Percent of respondents supporting job loss as a consequence for celebrating or justifying political violence by occupation type.

# If the following positions post on their personal social media account celebrating or justifying the killing of a political figure, should they lose their job?

political content online, we asked respondents "How often do you consider the following before posting (or declining to post) political opinions online?" including harassment, threats, or doxxing, job or professional consequences, social consequences, or legal consequences. As reported in Figure 7, roughly half of respondents did not feel this was a relevant question since they have no interest in posting about politics on social media. Of the remainder, respondents were divided. 53% worry about harassment very or somewhat often. 54% worry about professional consequences, while 58% and 47% worry about social and legal consequences, respectively. This amounts to roughly 1/4 of the Canadian public.

However, very few Canadians report suffering personal consequences from their speech. As described in Figure 8, 79% report

no consequences at all. 3–4% report suffering legal or professional consequences. The most frequently reported consequences were harassment and social conflict, which respectively 8% and 10% of Canadians say they experienced.

These findings broadly align with IU3, where about half of respondents reported no change in their political behavior due to concerns about political violence, while about one quarter indicated they would avoid posting opinions on social media, attending protests, or putting up political signs for that reason. The relatively low numbers of Canadians reporting actual negative consequences for their speech may not indicate the absence of risk, but could also reflect chilled speech, with some individuals avoiding expressing their political opinion out of concern for potential repercussions.

Figure 7.

How often respondents think of consequences before posting or failing to post political opinions online

**Figure 8. Percent of respondents experiencing consequences for expressing political opinions**

### **Conclusion**

Our findings suggest that an overwhelming majority of Canadians are opposed to all forms of political violence by people who share their political views, though it's troubling that this opposition softens somewhat towards harassment directed towards political candidates. They also mostly believe different partisan and ideological groups categorically reject political violence, though do tend to underestimate the extent to which political violence is rejected by their partisan or ideological opponents. This could be a driver of support for political violence or reflect projection effects where people predisposed towards violence project that support onto their political opponents.

Canadians are somewhat willing to trade off civil liberties to combat political violence but overall, they are somewhat divided. An overwhelming majority of Canadians believe there should be some consequences to speech that celebrates, justifies or encourages political violence – where most support legal prosecution or professional sanctions of those who incite violence (e.g. engage in acts that encourage violence). Support diminishes when the speech in question celebrates or merely justifies such violence.

At tension with these findings is the reality that the majority of Canadians believe freedom of expression in the country to be threatened to some degree. Notwithstanding this belief, only a minority of people report a reluctance to express themselves in online spaces because of possible consequences (though it's roughly half of people that actively engage in online political discussion). More importantly, extremely few Canadians report having experienced consequences from expressing themselves. For many Canadians, the threat perceived to freedom of expression appears to be rooted more in the fear of potential repercussions and self-censorship than in actual, lived experiences of consequences.

### **Methodology**

Below, we provide the wording of the vignettes used to measure support for different forms of political violence:

**Harassment:** Imagine that a person was convicted of harassment and uttering threats. This individual has political views similar to your own. He was arrested for repeatedly communicating threatening and hostile remarks to a political candidate from a political party you oppose over the period of a year on social media.

**Vandalism:** Imagine that a person was convicted of vandalism. This individual has political views similar to your own. He was arrested for smashing the windows of a local homeowner who was a known activist for a political cause you oppose. The action caused approximately \$2000 in property damage.

**Assault:** Imagine that a person was convicted of assault. This individual has political views similar to your own. He was arrested by police for throwing rocks at peaceful protesters in favour of a political cause you oppose. Although no one was seriously injured, paramedics bandaged a man with a head wound.

**Arson:** Imagine that a person was convicted of arson. This individual has political views similar to your own. He was arrested by police as he attempted to run from a fire he started at the local headquarters of an organization that supports a political cause you oppose. Although he waited for the building to close for the night, several adjacent buildings were still occupied.

**Assault with a deadly weapon:** Imagine that a person was convicted of assault with a deadly weapon. This individual has political views similar to your own. He was arrested by police after driving his car into a crowd of protesters in favour of a political cause you oppose. Although no one was killed, several individuals were seriously injured and one spent a month in the hospital.

### <span id="page-24-0"></span>**Incident Update 6**

# **Canadians' Awareness, Beliefs, and Reactions to the Charlie Kirk Shooting**

**Mathieu Lavigne, Chris Ross Media Ecosystem Observatory**

### **Key takeaways:**

- E **The assassination transformed Charlie Kirk from a wellknown internet figure to a widespread household name:** Prior to his assassination, 1 in 4 (28% of Canadians were familiar with who Charlie Kirk was, but almost all Canadians (88%) reported hearing about his death. Of those, nearly half (46%) identified social media as the largest source of information and discussions about the incident, with 67% of young Canadians aged 18–34 primarily hearing about his assassination from social media.
- E **Extreme reactions were overrepresented online:** More Canadians felt sadness and grief, condemned political violence, or expressed societal concern than they recalled seeing online. Meanwhile, happiness and outrage were more frequently seen online compared to how many Canadians personally felt those emotions.
- E **Most of the information Canadians flagged as false focused on the shooter's identity and motives, as well as Charlie Kirk's legacy:** Around a third (35%) of Canadians reported being exposed to false information about Charlie Kirk's death. False information they identified primarily concerned the shooter's identity and motives (39%), including his association with left ideology or the MAGA movement, and conflicting partisan interpretations of Kirk's stances and character (19%).
- E **Indicators of partisan polarization and chilled speech in Canada have remained stable over the course of 2025:** Partisan Canadians continue to express strong dislike toward opposing parties, and many Canadians remain uncomfortable sharing their political views online. However, our data show no evidence that these patterns intensified due to the assassination or other events in 2025.

### **Introduction**

### **Context**

The [assassination of U.S. political commentator Charlie Kirk](https://www.cdmrn.ca/publications/effects-of-the-charlie-kirk-assassination-in-the-canadian-information-ecosystem) sparked widespread discussion and elicited polarized reactions worldwide, raising questions about the nature of the information Canadians were exposed to and how such information affected public understanding and social cohesion.

In Updates [2](#page-50-0), [4](#page-40-0), and [5](#page-31-0), we identify the key narratives and reactions that circulated on X and TikTok around Charlie Kirk's assassination globally and in Canada. We found that the online environment was characterized by important divisions about the causes of political violence, the identity and motives of the killer, Charlie Kirk's legacy, and free speech. The nature of Canadian discussions paralleled that of the U.S., but their scope and visibility were comparatively less.

This update expands on these findings by evaluating the online information environment following Charlie Kirk's death through the eyes of a representative sample of Canadians. More specifically, we assess how widely Canadians heard about the incident and through which sources, the type of reactions and narratives they encountered online, and how these compare with their own feelings and perceptions, in order to help understand the interaction between the online information environment and Canadians' own perceptions of the event.

### **Key questions**

This incident update is particularly interested in evaluating the type of information that Canadians were exposed to and believed in regarding the Charlie Kirk assassination. We specifically focus on the following four questions:

- 1. How familiar were Canadians with Charlie Kirk, and to what extent did they hear about his assassination?
- 2. What emotions did the incident elicit among Canadians, and what emotions did they encounter or perceive in online discussions?
- 3. Which claims about the incident did Canadians encounter and consider false or misleading?
- 4. Have levels of partisan polarization and chilled speech in Canada increased since the beginning of 2025?

# **Approach and considerations**

This update primarily draws on data from an original survey conducted with a nationally representative sample of 1,459 Canadians between October 30 and November 7, 2025. Data is

<span id="page-25-0"></span>**Figure 1. Canadians' awareness of Charlie Kirk before his death by partisanship and age**

<span id="page-25-1"></span>**Figure 2. Canadians' awareness of the Charlie Kirk assassination and where they encountered discussion about it**

weighted by gender, age, and region according to the 2021 Census.

To measure affective polarization in the last section, we drew upon tracking survey data conducted monthly by the CDMRN from January to October 2025, including our election tracking survey fielded daily throughout the 2025 federal election. Each monthly survey includes 1400–1500 respondents and the election survey includes approximately 7,000 respondents from the campaign period.

### **1. How familiar were Canadians with Charlie Kirk, and to what extent have they heard about his assassination?**

Before his death, Charlie Kirk was known to a sizable minority of Canadians. As shown in [Figure 1](#page-25-0), 28% of Canadians knew who he was, and another 20% found his name familiar, while nearly half (49%) had never heard of him. There were modest partisan differences: Conservative supporters were somewhat more familiar with him (37%) than New Democrats (32%) or Liberals (26%). Younger Canadians stand out as the most likely to know who

<span id="page-26-0"></span>**Figure 3. Reactions to Charlie Kirk's death (focused on reactions reported by at least 5% of respondents)**

he was: 44% of those aged 18–34 knew who he was, compared with 25% of those aged 35–54 and 20% of those aged 55 and older.

Although relatively few Canadians knew of Charlie Kirk prior to his death, news of his death received widespread attention, with 88% reporting awareness of it. As a reference, 95% had heard about Donald Trump's assassination attempt in July 2024.

As [Figure 2](#page-25-1) illustrates, the incident received the most attention on social media, with nearly half of Canadians (46%) saying they experienced the most discussion there. In comparison, 28% said they primarily encountered discussion on television, 12% on news websites, and 9% through friends and family. The source of information exposure about the incident was strongly influenced by age: Canadians aged 18–34 were significantly more likely to hear about it on social media (67% compared to 24% among those aged 55 or older), while Canadians aged 55 or older were more likely to hear about it on television (47% compared to 8% among those aged 18–34).

In sum, while Charlie Kirk had some degree of online visibility in Canada prior to his death, particularly among younger generations, news of his assassination reached nearly all Canadians, and social media quickly emerged as the central space for public conversation about the event.

### **2. What emotions did the incident elicit among Canadians, and what emotions did they encounter or perceive in online discussions?**

To evaluate potential disconnects between online discourse and Canadians' own personal feelings, we asked respondents to report, in open-ended responses, the emotions they personally

felt upon hearing about Charlie Kirk's assassination (n=1,456), as well as the main emotions they observed in online discussions (n=1,454). A human coder and Claude Haiku 4.5 independently reviewed all responses to extract the most frequently mentioned emotional reactions, producing very similar results. We reconciled the outputs to form 14 categories based on three criteria: distinctiveness, comprehensiveness, and comparability between personal and online reactions. Finally, we used Claude Haiku 4.5 to code each response for the presence of these 14 reactions, allowing multiple reactions per response. For example, the following response was coded as "Yes" across three categories: "Sadness and grief," "Condemnation of violence," and "Societal concern."

*I was saddened by his death, not because he was a politician but because of his family left behind. People shouldn't be killed for doing their job, just like a doctor shouldn't be killed for doing his job, or a teacher etc. the world is a very scary place.*

Open-ended response to survey question asking what emotions they experience when they heard about Charlie Kirk's assassination

As shown in [Figure 3](#page-26-0), the primary emotion experienced by Canadians who heard about the assassination was sadness and grief (including sympathy for the family) (32%). About a quarter of respondents indicated that they didn't feel anything (23% reporting indifference or no emotion), many of whom specified they were unmoved because they didn't know who he was before his death. The third most commonly mentioned reactions involved condemnation of violence, and anger, disgust and outrage (both at 19%). A notable percentage of Canadians

<span id="page-27-0"></span>**Figure 4. False or misleading information identified by Canadians (open-ended)**

also expressed shock or disbelief (16%) and fear or concern for the state of (American) society and democracy (15%).

In contrast to the broad range of personal emotions expressed by our respondents, respondents recalled primarily extreme reactions in online conversations. Emotional reactions such as condemnation of violence, societal concern, and fear were rarely identified as common online reactions compared with how often Canadians personally felt them (4% vs. 19% for condemnation of violence; 7% vs. 15% for societal concern). Conversely, 29% of respondents identified anger, disgust, or outrage as prominent reactions online, yet only 19% personally experienced these emotions. Similarly, 15% reported seeing happiness or relief in online discussions, while just 5% felt these emotions themselves. These gaps likely result from a combination of factors: individuals with stronger views are likely to be more vocal and visible in online spaces; content that generates strong reactions tends to receive more engagement and consequently be more amplified by algorithms; and Canadians' were likely exposed to more emotionally charged reactions from U.S.-based users.

### **3. Which claims about the incident did Canadians encounter and consider false?**

Among Canadians who have heard about the shooting of Charlie Kirk, we evaluated the types of claims that Canadians were exposed to and perceived as false. 35% reported encountering information online they suspected was false or misleading (33% said they did not, and 32% were unsure). Those who reported exposure to false or misleading information were then asked to describe the content they had seen (n=435). We manually identified the most salient categories and classified these open-ended answers into four main categories, each containing subcategories:

**Shooter identity and motives:** Who is responsible for the death of Charlie Kirk and what their motivations were;

**Event:** How the event unfolded, whether Charlie Kirk is truly dead, and how the case has been handled;

**Kirk and family:** Information about Charlie Kirk and his wife (positions, legacy, reactions, etc.)

<span id="page-28-0"></span>**Figure 5. Canadians' belief in different explanations for the Charlie Kirk shooting**

**Criticism of information sources:** The role of various information sources (e.g., the Trump administration, mainstream media, social media, generative AI) in the spread of misinformation about the shooting.

[Figure 4](#page-27-0) presents the share of respondents who identified each of these false or misleading claims in their open-ended responses. As shown, 39% mentioned claims about the shooter's identity or motives. The most common claims included allegations that the shooter was from the "radical left" (8%); that the attack was carried out by the Trump administration, the MAGA movement, or someone close to Kirk (7%); that it was orchestrated by a foreign actor or Israel (3%); and that the shooter or his wife were transgender (1%).

Specific to the event, about one in five (18%) mentioned having seen false information about the assassination itself, including how the event unfolded (e.g., whether Kirk's children were present) (8%), the theory that Kirk is not actually dead (7%), and how the case has been handled (e.g., the wrong shooter being arrested) (4%).

Related to Charlie Kirk and his family, 22% mentioned information related to Charlie Kirk (19%) or his wife (3%). Those sympathetic to Kirk often suggested that left-leaning Canadians were taking his quotes out of context to imply that he supported positions he did not. Conversely, many Canadians argued that the Trump administration and the political right more broadly were trying to whitewash his reputation or portray him as a saint or martyr. Finally, Canadians criticized specific information sources for their role in spreading false information, especially the Trump administration and the DOJ (6%), AI-generated content (4%), the news (3%), and social media (2%).

To evaluate how widely these claims were believed in Canada, we asked respondents how likely they considered various explanations for Charlie Kirk's shooting. Canadians differ in which explanations for the assassination they find credible. As shown in [Figure 5](#page-28-0), a majority (69%) agree that the attack was politically motivated. About one in two Canadians indicated that it is "somewhat" or "very" likely that the attack was driven by online extremism and radicalisation (53%) or resulted mainly from mental health and personal issues (48%). Other narratives are believed by a small but visible minority of Canadians: 11% indicated that it is "somewhat" or "very" likely that the assassination was staged or faked, 14% believe it was coordinated or encouraged by foreign actors, and 25% believe it was part of a broader extremist plot involving multiple people. Notably, 14–25% of respondents selected 'don't know' for each explanation, indicating high uncertainty about the cause or motive of the assassination among the Canadian public.

### **4. Have levels of partisan polarization and chilled speech in Canada increased since the beginning of 2025?**

Our previous updates show that online discussions around the incident were highly polarized ([IU2](#page-50-0), [IU4](#page-40-0)) and that, in this polarized context, concerns around political violence may reduce

<span id="page-29-0"></span>**Figure 6. Affective polarization over time** 

**Figure 7. Comfort expressing political opinions online and offline over time** 

citizens' willingness to participate in the political conversation. In that context, we wanted to evaluate whether there had been any visible change in partisan polarization or chilled speech since the beginning of 2025.

Our monthly tracking survey measures Canadian attitudes towards political parties in Canada, a form of [affective polariz](https://www.mediatechdemocracy.com/all-work/mass-polarization-in-canada-whats-causing-it-why-should-we-care)[ation](https://www.mediatechdemocracy.com/all-work/mass-polarization-in-canada-whats-causing-it-why-should-we-care). We ask respondents whether they like or dislike parties on a scale from 0 to 100. [Figure 6](#page-29-0) shows averages for these at-

titudes across the major partisan groupings in Canada since January 2025. As could be expected, Canadians tend to have more positive feelings towards the party they identify with, with an average favourability in the mid-70s to low-80s. Liberal and NDP supporters tend to have negative feelings towards the Conservatives (in the low-20s) and relatively neutral feelings towards one another (around 50 on the 0–100 scale). Conversely, Conservative Party supporters express negative feelings towards both the Liberal Party and NDP (in the 20s).

[Research](https://www.mediatechdemocracy.com/all-work/mass-polarization-in-canada-whats-causing-it-why-should-we-care) has shown that this form of polarization3, that is to say the gap between in-group ratings and out-group ratings, has increased over time in Canada. Our tracking survey data show that although polarization remains high, it has been largely stable throughout 2025. The most notable movement is a slight improvement in feelings towards the Liberal Party following Mark Carney's selection as a leader, which has gradually reversed.

Our tracking survey also evaluates how comfortable Canadians feel expressing their political opinions online, responding to political content they disagree with, and expressing their political opinion at work or at school. The results, reported in Figure 7, indicate that a majority of Canadians do not feel comfortable expressing their political opinion, with comfort being slightly higher offline than online. This discomfort, while concerning, has remained largely stable throughout 2025.

### **Conclusion**

Our findings demonstrate that while Charlie Kirk was only known by a minority of Canadians, almost all of them have heard about his assassination. Canadians encountered the most discussion about the incident on social media, where the most contentious reactions were significantly more intense and polarized than the emotions Canadians reported feeling themselves. About a third of Canadians believe they were exposed to false or misleading information about the assassination, and primarily name the shooter's identity and motives, together with claims about Charlie Kirk's positions and character, among the false information they encountered. When asked whether they believe in different narratives about the incident, a majority of Canadians agree that the assassination was politically motivated and driven by online extremism, while narratives around foreign involvement or the attack being staged or faked were only believed by a minority of Canadians (11–14%). Given the nature of online discussions around the assassination and concerns around citizens' unwillingness to participate politically in a polarized information environment, we analyzed the evolution of partisan polarization and chilled speech since the beginning of 2025. We found that Canadians generally have very negative views towards parties on the other side of the political spectrum and do not feel comfortable posting their opinions online, but that these trends have not intensified in 2025. Overall, the results show that while online conversations around Charlie Kirk grew highly charged, Canadians' personal views remained more measured, illustrating the importance of distinguishing digital conversations from the attitudes of the broader Canadian public.

### <span id="page-31-0"></span>**Incident Update 5**

# **Mapping Controversial and Polarized Claims Following the Assassination of Charlie Kirk**

**Mika Desblancs-Patel, Esli Chan, Ben Steel, Mathieu Lavigne Media Ecosystem Observatory**

### **Key Takeaways**

- E **Most prominent claims contained a mix of speculation, blame, and critique:** Claims about the assassination of Charlie Kirk on X and TikTok primarily focused on partisan blame, critique of Kirk, and speculation about the shooter's identity and motive.
- E **Users were more likely to refute than to promote polarizing claims:** Across both X and TikTok, users predominantly directly refuted or offered competing narratives to polarizing claims rather than endorsing them. While X produced a higher volume of posts, TikTok drove substantially greater engagement.
- E **Polarization emerged from reactions to Kirk's death:** Claims related to the celebration of Kirk's death attracted the greatest volume of posts and engagement and showed the sharpest polarization. The conversation centered more on the incident's political ramifications than the circumstances of the death.
- E **Influencers lead the discussion in Canada**: Influencers were the most active voices discussing Charlie Kirk and received the most engagement in Canadian online spaces, particularly in conversations about public reactions and Kirk's legacy.

### **Introduction**

In this update, we aim to explore how polarized the conversation was following Kirk's death by identifying the most popular claims made following his assassination and evaluating the extent these claims were agreed with (supported), disagreed

with (refuted) or its themes discussed (other). To do so, we analyze the top claims that were made on X and TikTok in the aftermath of the shooting of Charlie Kirk on September 10th. More specifically, we evaluate the prevalence of these claims, the distribution of users' opinions about them, how they circulated online, and the role Canadian media, politicians, and influencers played in their dissemination.

In a previous incident update ([IU2](#page-50-0)), we used topic modelling methods to identify and group posts discussing similar themes. However, while thematic analysis of the incident helps us identify *what* people talked about, it does not distinguish trends in *how* users talk about an event. For example, [IU2](#page-50-0) found that users initially discussed the "transgender identity of the shooter" but did not distinguish between the argument that the shooter was trans, those claiming that trans identity was unrelated, and those referring to this narrative without making an explicit judgment about its truthfulness.

### **Key Questions**

We define a 'claim' as a meaningful and interesting statement or opinion that presents something as fact. We specifically look at the following questions:

- 1. What claims about Charlie Kirk's assassination circulated the most on X and TikTok?
- 2. How did these claims emerge?
- 3. Who circulated these claims in the Canadian information ecosystem, and how much engagement did they receive?

### **Approach & Considerations**

In a dataset of X and TikTok posts from September 5 to October 1, 2025, we extracted the most prominent claims by the number of posts count that advanced a narrative related to the death of Charlie Kirk, omitting generic statements (e.g., 'Charlie Kirk is dead') and manually identified claims that expressed an original stance or perspective on the incident (e.g., "Leftists are responsible for the murder of Charlie Kirk"). We then used a stance analysis model to evaluate how people engaged with these top claims in three main categories:

- (1) *supporting*: a post directly confirms or provides strong circumstantial evidence supporting a claim;
- (2) *refuting*: a text directly contradicts the claim or undermines the claim's plausibility without explicit denial;
- (3) *other*: a text is thematically related to the claim but the claim is neutral or the model doesn't express an opinion for or against it.

For example, suppose we had this post:

"Saw a clip of Charlie Kirk. That was horrific. Seen a lot of attention seeking lunatics on here celebrating his execution. Killed for having an opinion/debate and a lot of you sick individuals were cheering when it happened. He was a husband and a father."

We could evaluate the post against different claims and infer their relationship as follows:

- 1. People celebrated the death of Charlie Kirk **supporting**, the text's author explicitly states they saw users "celebrating" and "cheering".
- 2. Charlie Kirk's assassination was met with universal mourning—**refuting**, the text provides examples of individuals who did not mourn but instead cheered.
- 3. Left-wing individuals celebrated Kirk's death **other**, while the text refers to people celebrating and the fact that he died, the author does not identify their political affiliations.

By isolating specific statements to evaluate who initiates and amplifies them in the online ecosystem, we provide a granular characterization of how polarized and controversial statements emerge and influence political discourse.

### **1. What claims about Charlie Kirk's assassination circulated the most on X and TikTok?**

We identified the 17 most prominent and meaningful claims on X and TikTok related to the assassination of Charlie Kirk. We provide details about how claims were selected in the Methodology section below. We note that our collection efforts skew towards X — we collected 813,407 X posts and 10,937 TikTok videos to extract claims.

Most of these top claims broadly fall into six groups:

### *1. Public response to Kirk's death*

Public discourse following his death yielded top claims about what the appropriate response should be, extending to cultural conversations about public behavior:

*"People celebrated that Charlie Kirk died"*

*"People should get fired for celebrating Charlie Kirk's death"*

*"Jimmy Kimmel's comments about Charlie Kirk were inappropriate or incendiary"*

### *2. Partisan critique*

### *a. Critique of the left*

Users deliberated the responsibility of the ideological left for Kirk's death and political violence, their reactions to his death, and how much they care about free speech. These statements include:

*"Leftists are responsible for the murder of Charlie Kirk"*

*"The Democrats are responsible for the murder of Charlie Kirk"*

*"Leftists are responsible for political violence"*

*"Left-wing individuals celebrated the fact that Charlie Kirk died"*

*"The left does not care about free speech"*

### *b. Critique of the right*

In contrast, discourse about the ideological right was limited within the top 17 claims with only one prominent claim focused on free speech:

*"The right does not care about free speech"*

### *3. Speculation about the shooter*

Speculation around the shooter's connection to foreign actors, identities, and ideologies demonstrated the public's desire to understand the underlying motive of the assassination. Claims included:

*"Israel was involved in the shooting of Charlie Kirk"*

*"The Shooter or Tyler Robinson was radicalized by the left"*

*"The Shooter or Tyler Robinson was radicalized by trans ideology"*

*"The shooter or Tyler Robinson is a Groyper"* 

*"Trans ideology was part of the murder"* 

### *4. Criticism of Charlie Kirk*

A large number of users discussed the character and legacy of Kirk in relation to the claim:

*"Charlie Kirk was a racist and a bigot"*

### *5. Schools are indoctrinating students*

In line with partisan critiques of the left, many users accused universities (and schools more broadly) of indoctrinating students with left-wing ideas. Some users argued it was these ideas the shooter used to justify his actions. The main claim discussed was:

*"Schools are indoctrinating students"*

<span id="page-33-0"></span>**Figure 1. Sum of likes across posts on X and TikTok supporting each claim, labelled by individual claim, its total number of likes, and coloured by claim group.**

### *6. Antifa should be declared a terrorist organization*

After the assassination, many users online linked the shooter to ANTIFA and, in late September, President Trump announced an Executive Order to have the organization designated as a terrorist organization. Those discussions are reflected in the claim:

*"Antifa should be declared a terrorist organization"*

We note that the above should not be taken to mean they were the only claims made specifically about Charlie Kirk and the assassination, but rather the ones that were most consistently identified in our data, which is sensitive to the consistency of the language used when users make a statement.

To assess how widely each claim circulated on X and TikTok, [Figure 1](#page-33-0) reports the total number of likes on posts supportive of each claim, with colors distinguishing the six claim groups. Table 1 provides a broader view of the discussion by reporting, for each claim, the total number of posts and likes per stance (support, refute, other). Overall, we find that *other* stances on posts largely outnumbered both *supporting* and *refuting* posts for most claims. We take this to reflect both the high number of posts in which users did not clearly state a position, the difficulty our model faced in detecting implicit stances in context-dependent social media data, but also the thematic uniformity of our dataset; all posts were about Kirk's death, and the ensuing online conversation. Nevertheless, we identify three key points:

### *1. X produces narratives while TikTok popularizes them*

While we found 12 times more posts related to the claims on X than TikTok (~133,000 vs. ~10,000), TikTok tallies up about 3.5 times more likes than X (~478,000 vs. ~133,000). In other words, narratives (posts) appear to originate on X, but engagement

| Claim                                                                              | Count Twitter (left) & TikTok (right) |      |         |     |        |      |       | Likes Twitter (left) & TikTok (right) |        |        |         |        |        |        |        |        |
|------------------------------------------------------------------------------------|---------------------------------------|------|---------|-----|--------|------|-------|---------------------------------------|--------|--------|---------|--------|--------|--------|--------|--------|
|                                                                                    | Other                                 |      | Support |     | Refute |      | Total |                                       | Other  |        | Support |        | Refute |        | Total  |        |
| The shooter or Tyler Robinson was radicalized by trans ideology                    | 380                                   | 31   | 203     | 8   | 365    | 22   | 948   | 61                                    | 89.9K  | 1.2M   | 40.4K   | 20.2K  | 135.8K | 924.4K | 266.1K | 2.2M   |
| The right does not care about free speech                                          | 170                                   | 88   | 39      | 0   | 1.7K   | 105  | 1.9K  | 193                                   | 22.0K  | 5.0M   | 46.0K   | 0      | 1.3M   | 1.9M   | 1.4M   | 6.8M   |
| Leftists are responsible for the mur-<br>der of Charlie Kirk.                      | 634                                   | 181  | 389     | 1   | 9.3K   | 190  | 10.3K | 372                                   | 480.3K | 5.1M   | 342.7K  | 16.4K  | 7.5M   | 5.0M   | 8.3M   | 10.1M  |
| The shooter or Tyler Robinson is a<br>Groyper                                      | 428                                   | 137  | 102     | 4   | 66     | 6    | 596   | 147                                   | 140.6K | 3.6M   | 28.0K   | 75.4K  | 6.1K   | 314.7K | 174.7K | 4.0M   |
| Schools are indoctrinating students                                                | 18                                    | 5    | 20      | 1   | 0      | 0    | 38    | 6                                     | 495    | 127.5K | 24.5K   | 18.1K  | 0      | 0      | 25.0K  | 145.6K |
| People celebrated the fact that<br>Charlie Kirk died.                              | 1.9K                                  | 2.0K | 14.8K   | 527 | 29.5K  | 1.2K | 46.2K | 3.8K                                  | 1.3M   | 106.7M | 19.8M   | 17.7M  | 29.2M  | 66.7M  | 50.3M  | 191.1M |
| Charlie Kirk was a racist and a bigot.                                             | 1.8K                                  | 728  | 596     | 25  | 1.2K   | 16   | 3.5K  | 769                                   | 1.4M   | 32.0M  | 1.1M    | 925.7K | 2.9M   | 968.3K | 5.3M   | 33.9M  |
| Left-wing individuals celebrated<br>the fact that Charlie Kirk died.               | 20.1K                                 | 2.7K | 6.4K    | 131 | 7.2K   | 146  | 33.6K | 3.0K                                  | 19.2M  | 122.8M | 8.1M    | 2.6M   | 5.2M   | 6.4M   | 32.5M  | 131.8M |
| Trans ideology was part of the mur-<br>der                                         | 133                                   | 7    | 49      | 5   | 257    | 10   | 439   | 22                                    | 109.3K | 258.1K | 17.1K   | 12.2K  | 98.1K  | 464.0K | 224.4K | 734.3K |
| The Left does not care about free speech.                                          | 129                                   | 18   | 54      | 0   | 94     | 1    | 277   | 19                                    | 25.6K  | 698.9K | 4.1K    | 0      | 8.1K   | 2.4K   | 37.7K  | 701.3K |
| ANTIFA should be declared a terrorist organization.                                | 88                                    | 10   | 5       | 0   | 9      | 0    | 102   | 10                                    | 9.2K   | 30.3K  | 7       | 0      | 243    | 0      | 9.4K   | 30.3K  |
| Israel was involved in the shooting of Charlie Kirk                                | 28                                    | 11   | 10      | 0   | 2.6K   | 40   | 2.7K  | 51                                    | 3.7K   | 489.6K | 25      | 0      | 1.9M   | 930.4K | 1.9M   | 1.4M   |
| The Democrats are responsible for<br>the murder of Charlie Kirk.                   | 353                                   | 64   | 10      | 0   | 5.6K   | 70   | 6.0K  | 134                                   | 63.4K  | 1.5M   | 1.2K    | 0      | 2.6M   | 2.2M   | 2.6M   | 3.8M   |
| People should get fired for celebrating Charlie Kirk's death                       | 20.8K                                 | 1.6K | 1.0K    | 20  | 611    | 9    | 22.5K | 1.6K                                  | 20.9M  | 78.8M  | 1.2M    | 245.6K | 155.1K | 1.4M   | 22.3M  | 80.5M  |
| Leftists are responsible for political violence.                                   | 178                                   | 30   | 136     | 1   | 1.1K   | 45   | 1.4K  | 76                                    | 36.1K  | 859.0K | 192.5K  | 830    | 1.1M   | 1.4M   | 1.3M   | 2.2M   |
| The shooter or Tyler Robinson was radicalized by the left.                         | 274                                   | 41   | 644     | 14  | 498    | 39   | 1.4K  | 94                                    | 190.6K | 1.2M   | 269.5K  | 381.5K | 175.4K | 1.8M   | 635.5K | 3.4M   |
| Jimmy Kimmel's comments about<br>Charlie Kirk were inappropriate or<br>incendiary. | 301                                   | 138  | 809     | 42  | 87     | 4    | 1.2K  | 184                                   | 62.1K  | 3.3M   | 700.1K  | 2.0M   | 23.2K  | 16.2K  | 785.4K | 5.2M   |

<span id="page-34-0"></span>**Table 1. Count of posts and likes per stance for each claim**

is substantially higher on TikTok. This pattern is even more stark for posts expressing a support or refute stance: across all claims, there were 30 times more X posts (~85,000 vs. ~2,700), but TikTok posts generated 30% more likes than X (~115 million vs ~85 million). Taken together, these findings suggest that while X serves as a major site of content production, TikTok is the primary site of amplification, with narratives gaining far more visibility and impact once they reach TikTok.

### *2. Celebrations of Kirk's assassination was a driver of polarization*

As seen in [Figure 1](#page-33-0), claims asserting that people are celebrating Kirk's assassination garnered the most interest overall. They represented by far the most supported type of claims by engagement and exhibited highly polarized debate, as highlighted by the high volume and engagement of posts both supporting and refuting celebration claims. Supportive posts generally accused people of celebrating Kirk's death, while refuting posts emphasized mourning. This suggests that debates over the aftermath and public reactions to Kirk's assassination received more engagement than accusations about the motive or causes

of the incident. Users were interested in the potential fallout and emergence of socio-political conflict in wider society, rather than focusing solely on the incident itself.

### *2. Overall, users refuted claims*

Table 1 shows that across both platforms, the number of posts refuting claims far exceeded the number supporting them. In total, posts on X refuting claims received 52 million likes (out of 85 million), while those on TikTok received 90 million likes (out of 115 million). For example, X and TikTok users largely refuted the claim that leftists and democrats were responsible for the shooting, as shown by the volume of refute posts for claims associated with partisan critiques of the left in [Table 1](#page-34-0). This suggests that left-leaning users were providing competing theories, or actively pushing back against narratives blaming them for the assassination and the climate of political violence in the United States. Likewise, there were disproportionately more refuting claims for posts critical of Kirk and about the ideological right's lack of care for free speech. This demonstrates that right-leaning users praised Kirk and countered narratives that disparaged him or the right.

### **2. How did these claims emerge?**

In this section, we evaluate the first and most popular posts that emerged on X and TikTok related to three particular claims related to the identity and motives of the shooter: 1) Trans ideology was part of the murder; 2) The shooter or Tyler Robinson was radicalized by leftist ideology; and 3) The shooter or Tyler Robinson is a Groyper. We evaluate these claims as we found them to be particularly polarizing. Unlike in the previous section where we automated the discovery of posts supporting, refuting, or related to a claim, for this task, we manually assess our dataset of X posts and TikTok videos to ensure we are able to find the earliest post possible. Evaluating the first and most popular post reveals how claims emerge and drive speculation online. In this context, we note that the death of Charlie Kirk was recorded on September 10, 2025 12:23 MDT and the identification of the shooter was on September 12, 2025, 8:13 MDT.

### *Claim 1. "Trans ideology was part of the murder"***:**

The first post on X we were able to find drawing a link between transgender people and the shooting was made on September 10th at 12:41:59 MDT, just under twenty minutes after Charlie Kirk was shot. The post came from a verified crypto account which garnered over 50k views, and stated:

"Charlie Kirk having a speech about whether trans people should get their guns taken away, and he gets shot by a trans. cant make this up"

The most popular post came from a prominent right-wing influencer on September 13th, a day after Tyler Robinson was revealed as the primary suspect in the assassination. Their post garnered over 156,000 likes and 4.4 million views; they wrote:

"I've independently confirmed that Charlie Kirk's assassin did in fact have a transgender boyfriend. This is LGBT extremism. A lot more will be revealed soon. The "far right" narrative is absolute bullshit. Not even close to true. Don't fall for it."

On TikTok, the initial posts drawing this link were posted 24 hours later (Sept. 14th, 2025), reacting to a later amended [Wall](https://www.wsj.com/livecoverage/charlie-kirk-shot/card/ammunition-in-kirk-shooting-engraved-with-transgender-antifascist-ideology-sources-pdymd1sXXMSlVRhpvR4b?mod=e2tw)  [Street Journal article](https://www.wsj.com/livecoverage/charlie-kirk-shot/card/ammunition-in-kirk-shooting-engraved-with-transgender-antifascist-ideology-sources-pdymd1sXXMSlVRhpvR4b?mod=e2tw) which had falsely claimed the FBI had found "transgender and antifascist ideology" on the bullet encasings. We found posts by influencers along [news organizations](https://www.tiktok.com/@abc7la/video/7548906776159145229)  covering the story which were viewed 128,000 and 139,000 times.

However, unlike on X, these TikTok videos did not push the narrative as much as they simply reacted to what seemed like a credible news story at the time. In fact, we found that both early and popular videos on TikTok mostly reported alleged FBI discovery rather than the claim itself, suggesting that TikTok was not an early, nor main driver for the narrative.

### **Claim 2.** *"The shooter or Tyler Robinson was radicalized by leftist ideology"***:**

On X, the most popular and first post arguing Robinson's affiliation with leftist ideology as a radicalizing force was posted on September 10th at 13:07 (44 minutes after Charlie Kirk's death) by a verified account with a large following, but no affiliation to any media or political organization. This post garnered over 28 million views and 63k likes:

"The left-wing mainstream media, as well as figures like Gavin Newson claiming Trump plans to be dictator, has created a climate of hysteria against right-wing figures that could radicalize any number of unstable people to engage in political violence. Pray Charlie Kirk is ok."

On TikTok, the first post we were able to find was only posted on September 10th at 19:18 MDT (just under 7 hours after Kirk's death), and garnered over 2.9 million views. The claim within the video stated:

"one of these psycho fucking leftists decided to try to kill him"

In contrast to the most popular post on X, the most popular TikTok videos related to this claim were either posts by news organizations relaying official government narratives or users arguing against the narrative that he was radicalized by leftist ideology. For example, one of the most popular videos (3.6 million views) was from a political influencer who argued:

"this was another example of another shooting from a young white male from a conservative pro-gun religious background"

before also stating that it is hypocritical and false to blame left-wing thought for the assassination.

### **Claim 3. "***The shooter or Tyler Robinson is a Groyper".*

On X, the first post related to the claim that the shooter is a Groyper was identified on September 10th at 12:48 MDT, just 20 minutes after the shooting. Although it did not identify Robinson, it discussed the association or intent of the shooter. The claim was posted by a verified account and received 348,000 views; it stated:

"So was Charlie Kirk shot by a leftist or a groyper? Equally likely at this point"

One of the most popular posts was on Sept. 12th at 9:04 MDT by a popular political commentator after more information about the shooter became available. The post received 6,000 likes and had over 3.4 million views. It stated:

So the fake claims of "trans" and "antifa" symbols comes from both ammo stamps and casing inscriptions wrongly reported.

**Figure 2. Total likes on X and TikTok for posts supporting or refuting each claim, broken down by user type (Left) and stance (Right)**

They try to paint Tyler as left, but like Crooks this is a fringe right wing Groyper.

Here's the details on registered Republican, Tyler Robinson"

On TikTok, the first post we identified tentatively pushing the claim was on September 12th 11:00MDT and received 63,000 views. It claimed:

"Now that it looks like it's a groyper"

The most popular post came days later on September 15th 11:29 MDT by an online political and cultural influencer arguing that:

"they're not going to release [Tyler Robinson]'s internet history because it will point directly to the fact that he is a groyper"

Most of these posts originate from individual verified accounts with significant followings, political commentators or influencers, or news organizations. This demonstrates how emerging narratives on these platforms are largely driven by high-visibility actors who steer discourse. These examples illustrate how fast conspiratorial and polarized claims can emerge immediately following an event, with some narratives appearing before factual information has been fully established.

### **3. Who circulated these claims in the Canadian information ecosystem, and how much engagement did they receive?**

In this section, we assess the impact of these claims in the Canadian information environment. To evaluate how Canadians circulated and engaged with these claims, we collected all posts on X (excluding retweets) and TikTok transcripts containing the word "kirk" in our seedlist of influential Canadian political figures (news outlets, influencers, politicians, and civil society organisations) between September 5th and October 1st, 2025. In total, we found 3,267 posts on X and 682 posts on TikTok for a total of 3,949 posts. We then evaluated the stance of each post with respect to the 17 claims we had found above to evaluate their prevalence and the engagement they received within the Canadian information ecosystem.

In Figure 2,right side, we show the total number of likes on X and TikTok for posts supporting or refuting each of the 17 claims. The left panel breaks down likes by the type of user who posted them, while the right panel shows the sum of likes by stance. Since our dataset contained only 19 posts from civil society organizations that received little engagement, we omitted to plot them. We further note that three claims, "Israel was involved in the shooting of Charlie Kirk," "The Shooter or Tyler Robinson is a Groyper," and "ANTIFA should be declared a terrorist organization," received minimal engagement, which is why they are not clearly visible in the figure.

As shown in Figure 2, for posts taking a supporting or refuting stance on the selected claims, Canadians primarily engaged with influencer content following the assassination of Charlie Kirk. Influencers received 89% likes on posts supporting or refuting a claim, while news outlets only received 7.5% and politicians barely 3%, even though news outlets produced nearly half as many posts as influencers (233 vs 451). We conclude that influencers were the main drivers of opinionated conversation around the assassination of Charlie Kirk in the Canadian information ecosystem. This demonstrates how influencers played a key role in shaping how narratives were framed and subsequently consumed. Consistent with our [recent brief on the](https://meo.ca/work/power-shift-the-rise-of-political-influencers-in-canada)  [political influence of Canadian influencers](https://meo.ca/work/power-shift-the-rise-of-political-influencers-in-canada) and [IU2](#page-50-0), they provide interpretive frames that drive Canadians' understanding of political events.

Similarly to the general conversation we studied in the initial part of this update, claims regarding users celebrating the assassination of Charlie Kirk received the most engagement4, a little over 1 million likes. We also find that comments about Kirk's character were quite popular: the claim regarding Kirk being a racist and a bigot generated the second most engagement, with influencers mostly refuting the claim. Claims blaming the left for the assassination and for political violence also received comparatively high engagement, most of which came from supportive posts. Interestingly, while posts *refuting* polarizing claims received most of the engagement in the overall online environment, posts *supporting* these claims received more engagement when focusing on Canadian political figures (news outlets, influencers, politicians, and civil society organisations). While our data do not allow formal testing, this pattern may reflect either the greater supply of content from right-leaning

influencers in the Canadian context (most top claims are congenial to the right) or the fact that much of the refuting content originated from ordinary users rather than opinion leaders. Finally, we find that posts making a stance on Israel's role in the shooting, the shooter's affiliation to the Groyper movement, and the status of ANTIFA as a terrorist organization received little to no attention among Canadian influencers.

The high concentration of engagement on influencer content in Figure 3 does not mean, however, that news outlets had no influence on the conversation. When we also consider posts discussing each claim, not just those that supported or refuted them, we find that news outlets received 30% likes. Similar to [IU2](#page-50-0), news outlets largely played an informative role, providing updates on the public's response and the investigation as information became publicly available.

Finally, politicians' behaviour also aligned with our findings in [IU2](#page-50-0): they largely remained out of the conversation, providing little to no commentary whether promoting or discussing different narratives. Civil society organizations similarly produced very little content supporting or refuting these claims, and the posts they did publish received limited engagement.

These findings suggest that American political events gain traction in the Canadian information ecosystem but do not necessarily dominate Canadian online discourse. Canadian influencers play a key role in interpreting events and disseminating claims and narratives in the Canadian context, whereas news outlets see their role as providing informational updates on ongoing events. The Kirk incident, however, does not reveal a pronounced partisan debate among Canadian politicians, as shown by the minimal engagement and commentary from this group.

### **Conclusion**

Our analysis demonstrates how X and TikTok users engaged with polarizing and conspiratorial claims about the assassination of Charlie Kirk. Top claims about Kirk demonstrate a wide range of partisan discourse, speculation around the shooter and his motives, the role of political ideologies in the assassination, and critique of Kirk's character. These claims generated greater posting activity on X, but most of the engagement occurred on TikTok. Our claim-extraction model suggests that a large bulk of the discourse focused on refuting partisan blame. This pattern is largely consistent across posting and engagement activity.

The claims responsible for the largest number of posts and engagement focused on reactions to Charlie Kirk's death, such as "People celebrated the fact that Charlie Kirk died". The high prevalence of both *refuting* and *supporting* content demonstrates that there was no clear ideological or narrative dominance, highlighting the polarized nature of online discussion.

Our analysis indicates that speculative claims about the identity and motive of the shooter emerged rapidly following the incident, especially on X. Polarized and conspiratorial narratives that originated in the United States permeated the Canadian online environment. Canadian influencers were drivers of these narratives, but overall activity was limited.

Overall, this incident update illustrates how major political events can quickly spark polarized, conspiratorial narratives in the online environment. Audiences are quick to engage with these narratives, with discussions reflecting disagreements over the causes of the incident and its broader social consequences. The influence of American events in Canadian online discourse underscores the permeability of the Canadian information environment and the importance of monitoring how U.S.-based polarization can transport into the Canadian context.

### **Methodology**

### **What do we mean by claims?**

We understand a 'claim' to be a statement provides a meaningful and interesting stance, framing an opinion about a particular viewpoint or issue as fact or truth. We view the process of differentiating claims as extracting all atomic units within a social media post or transcript that express a single stated opinion or assertion about the world. For example, on September 18, former US President Barack Obama posted on X:

*"After years of complaining about cancel culture, the current administration has taken it to a new and dangerous level by routinely threatening regulatory action against media companies unless they muzzle or fire reporters and commentators it doesn't like."*

This post can be broken down into the following claims:

- 1. The current administration has complained about cancel culture for years.
- 2. The current administration has taken cancel culture to a new and dangerous level.
- 3. The current administration routinely threatens regulatory action against media companies.
- 4. The current administration threatens regulatory action against media companies unless they muzzle or fire reporters and commentators it doesn't like.

### **Data collection**

In this update, posts were collected for both X (Twitter) and TikTok.

For X, we collect posts from a large collection of keywords related to the assassination between the 5th of September and 1st of October, 2025. Filtering was done in two steps. First, we leveraged Twitter's semantic and keyword search capabilities by keeping all posts resulting from searching: "@charliekirk11", "Charlie Kirk", "Kirk", "Turning Point", "TPUSA", "Tyler Robinson", "Charliesmurderers" and "Erika Kirk". Second, for all other tweets, we kept those containing the keywords: "kirk", "tyler", "robinson", "tpusa", "turning point". Finally, we further filtered by keeping only posts that weren't replies to other posts.

For TikTok, because the search functionality does not return a large volume of recent posts, we used a snowball sampling method. We started by collecting posts returned with the keyword "Charlie Kirk". Then, for each video, we collected the top 20 recommended videos, continually looping this step to obtain more videos. Once we collected the list of videos, we transcribed them all.

In total, we collected 813,407 X posts and 10,937 TikTok videos.

### **Claim extraction**

We used a small LLM (Qwen/Qwen3-8B) and used a few-shot prompt to extract every statement made in an X post or TikTok transcript. If no statements or claims are found, the model is instructed to return an empty list of claims. In total, we were able to extract 2,943,121 claims on X, and 162,712 claims on TikTok.

### **Embeddings**

We used Qwen/Qwen3-Embedding-0.6B to generate embeddings for each claim. The motivation behind generating the embeddings of the claims instead of the entire text is to avoid drowning out claims that might occur in multiple posts or transcripts with surrounding text that might contain different or unrelated content. This process is especially important for longer texts like TikTok transcripts.

### **Clustering**

We clustered the claims using HDBSCAN for both X and Tik-Tok. Due to the difference in volume between TikTok videos and X posts, we set the minimum cluster size at 50 and 150, respectively. As we were interested in getting the most widely circulated claims, the larger minimum cluster size pushes the algorithm to find large groups of claims.

### **Canonical claim generation**

After clustering, we use BERTopic's default [cluster naming meth](https://maartengr.github.io/BERTopic/getting_started/vectorizers/vectorizers.html)[od](https://maartengr.github.io/BERTopic/getting_started/vectorizers/vectorizers.html) to generate a keyword representation for each cluster. Then, for each cluster, we extracted representative exemplars by 1) finding the top 20 most similar claims for each cluster to their keyword by cosine similarity, and 2) sampling 10 claims using [toponymy's saturated coverage method](https://toponymy.readthedocs.io/en/latest/exemplar_texts.html). We found that simply using cosine similarity to the cluster keyword representation tended to be sensitive to certain tokens and used an alternate method to enhance the breadth of our exemplar selection. Finally, once we generated the 30 representative exemplars, we leveraged another small LLM (Qwen/Qwen3-14B) to generate new representative claims, which we call "canonical claims", by synthesizing the dominant claims among the exemplars.

### **Claim selection**

One issue with extracting claims is that, often, the most popular claims are generic, commonly used phrases or claims that do not express a meaningful or interesting stance. For example, one of the top claims was "Kash Patel is the director of the FBI", or "Charlie Kirk is dead". To curb this issue, we manually reviewed the top 50 clusters and selected representative claims that present an interesting perspective on the incident. Overall, we found that X and TikTok posts shared the same set of interesting claims, so we constructed a conjoint list across both platforms. Our final list of claims contained 17 individual claims. While we had hoped to consider more in our final list, we had to limit our selection due to the high cost of using Claude Haiku 4.5 for the entailment process (described below).

### **Entailment**

In addition to identifying the most circulated claims, we also wanted to evaluate how Canadians felt about these claims, i.e. the extent to which Canadians were actively *promoting* or *refuting* certain claims rather than simply mentioning them.

To do so, we constructed post-claim pairs by selecting posts whose semantic similarity to the claim exceeded 0.7. Finally, we leveraged Claude's Haiku 4.5 to classify a post-claim relation into seven categories of stances:

- (1) *supporting*—text directly confirms or provides strong circumstantial evidence supporting a claim;
- (2) *refuting*—text directly contradicts the claim or undermines the claim's plausibility without explicit denial;
- (3) *other*—the text is thematically related to the claim but the claim is neutral or the model does not determine if it expresses an opinion for or against it
- (4) *irrelevant*—there is no meaningful connection between the post and the claim

### **How is a claim different from a topic?**

As highlighted earlier, the aim of this incident update and a claim-extraction approach is to provide a description of which narratives individuals *supported* or *refuted,* rather than evaluate the *types* themes discussed. In light of this, our approach differs in two key aspects. First, while topic modelling finds similarity between posts by grouping together documents whose embeddings are similar, our approach groups together individual claims that are made within each document. This is especially helpful when working with long documents (e.g. TikTok transcripts or long X posts) where multiple narratives are being pushed. While an embedding model will struggle to capture all this information within one vector representation, breaking a post up into claims helps us find similarities between parts of a document. Secondly, our use of a claim entailment model makes it possible to differentiate between posts who both talk about something similar (e.g. Antifa and terrorism), and those who make a claim about Antifa and terrorism (e.g. Antifa is a terrorist organization).

### <span id="page-40-0"></span>**Incident Update 4**

# **Shockwaves on TikTok: Understanding the International Reaction to the Assassination of Charlie Kirk**

**Benjamin Steel, Diya Jiang, Mathieu Lavigne Media Ecosystem Observatory**

### **Key takeaways**

- E **Conversation went global:** Discussion about Kirk's death went far beyond American borders. In total, users outside the United States produced 42% of videos, and received 42% of the total views and 33% of likes.
- E **Engagement was highest in English-speaking countries:**  While U.S. creators dominated conversation, users in other English-speaking countries, particularly Canada and Australia, posted at similar rates relative to their population size.
- E **Political violence was condemned globally:** While the incident generated polarized debates about Charlie Kirk's legacy and the U.S. political context, content creators globally expressed predominantly critical reactions to political violence, with only 1% expressing support.

### **Introduction**

The [assassination of Charlie Kirk](https://www.cdmrn.ca/publications/effects-of-the-charlie-kirk-assassination-in-the-canadian-information-ecosystem) triggered a vast social media discussion both in the United States and internationally. As the U.S.'s polarized context appears to increasingly shape political debates abroad, this event provides a window into how online conversations unfold after instances of political violence, shedding light on both the transformation of existing tensions and the emergence of new ones.

In [Update 2](#page-50-0), we conducted a topic modelling of the key narratives present on X in Canada and internationally. We found that initial discussions were driven by conspiratorial and partisan narratives, before shifting toward sustained debates over political violence and free speech. Canadian discourses mirrored those in the U.S. but remained limited in volume and attention.

In this update, we expand on [Update 2](#page-50-0) by examining the global reaction to Charlie Kirk's assassination on TikTok. This platform is particularly relevant for studying the global conversation because its content-driven algorithm, which is less reliant on pre-existing social networks, encourages otherwise unconnected, multilingual audiences to engage in the same conversation. We focus on this platform to identify the countries where discussion of Kirk's assassination was most widespread and to examine how content creators in these countries positioned themselves around three central themes in the discussion: Charlie Kirk and his legacy, Donald Trump, and political violence in the U.S. This analysis contributes to a deeper understanding of the international echoes of U.S political shocks and how they shape global conversations about political violence.

### **Key questions**

This update aims to answer the following two questions:

- 1. Which countries generated the most discussion about the event?
- 2. How did reactions to the event vary between countries?

# **Approach and considerations**

To answer these questions, we collected TikTok videos using a two-step approach: first, by searching for posts containing the terms "charlie" and "kirk" in their descriptions, and second, by collecting the posts related to these videos (i.e., *Related content*). In total, we collected 7,465 videos in the weeks following the assassination (Sept. 12–Nov. 4), with the post creation times ranging from the day of the event (Sept. 10) up to Nov. 4.

To analyze the content of the videos, we used fine-tuned large language models (LLMs) and other computational techniques to find the main issues of discussion, and stances on those issues, in the video descriptions and transcripts. This classification involved a small and known error. For this reason, we excluded videos using languages where the models perform poorly (e.g., Vietnamese).

Given resource constraints, this update's data collection was not exhaustive and does not capture the full universe of TikTok posts about the assassination. Our approach relied on searches and snowball retrieval of related posts, which tend to prioritize content with higher visibility, stronger engagement signals, or greater algorithmic reach. Although this method does not retrieve every relevant post, it is appropriate for identifying the

<span id="page-41-0"></span>**Figure 1. Top 5 countries of origin of videos on the subject of Charlie Kirk, sorted by absolute number of posts.**

<span id="page-41-1"></span>**Figure 2. Top 5 countries of origin of videos on the subject of Charlie Kirk, sorted by number of posts relative to the population of the country.**

dominant narratives, most circulated reactions, and cross-national patterns that shaped discussions on TikTok. As such, the dataset offers a strong representation of the most prominent narratives and reactions surrounding the event.

### **1. Which countries generated the most discussion about the event?**

We started by examining how many videos users from each country posted in response to the Charlie Kirk assassination, as well as the engagement they received in terms of views and likes. [Figure 1](#page-41-0) shows the total number of videos posted by the top five countries, with all remaining countries grouped in an "Other" category. Although the top-five selection is based on production volume, the figure also reports the views and likes these videos received. While these plots depict that the majority of posts emerged in the U.S. (where the assassination took place), they also demonstrate that the conversation extended far beyond U.S. borders. Canada, the United Kingdom, Brazil, and Australia stood out for both the number of posts they produced and the engagement those posts received, with many other countries contributing smaller but noticeable shares. In total, users outside the United States produced 42% of videos, and received 42% of the total views and 33% of likes.

However, these raw numbers can be skewed by the population size of each country. To account for this, [Figure 2](#page-41-1) shows the number of posts relative to a country's population size. Here, another pattern emerges: on a per-capita basis, users from other English-speaking countries, especially Australia and Canada, posted at a comparable rate to those in the United States.

We hypothesize that we observe these trends for a variety of reasons. Australia's high engagement may be explained by the success of Turning Point Australia, a youth-oriented conservative organization modeled after Charlie Kirk's Turning Point USA. Founded in 2019 by Australian commentator Joel Jammal, the organization gained visibility and developed ties with Australian conservative figures. While Kirk was not formally the lead of the Australian branch, he frequently [commented](https://www.news.com.au/national/politics/conservative-activist-charlie-kirk-criticised-australian-politics-for-years-before-death/news-story/7e4ce2f310b206b197d6a7051fac2860) on

<span id="page-42-0"></span>**Figure 3. Breakdown of stance on three prominent topics related to Charlie Kirk's assassination.** 

Australian politics, criticizing government policies on issues such as free speech, COVID-19 restrictions, and immigration. Canada's high posting rate likely reflects its geographic and cultural proximity to the U.S., as well as the interconnectivity between their information ecosystems, which exposes Canadians to U.S. political media and influencers. Shared platforms and a high level of cross-border media consumption mean that Canadian users are consistently exposed to American political narratives, especially those involving high-profile political elites. Despite geographic distance, the UK and New Zealand also demonstrated high engagement, probably due to their shared anglophone information environments.

In the middle and right panels of [Figure 2](#page-41-1), we divide the views and likes of videos from each country by that country's population. Although online audiences are highly globalized, this measure provides a rough proxy for how the content performed relative to its 'natural' home audience. In general, engagement per capita aligns with posting volume, although posts from the United Kingdom received more engagement and those from Canada received less engagement, despite their posting volumes. These patterns likely reflect variation in the global reach of creators who posted about Charlie Kirk.

Overall, these patterns demonstrate that while the conversation around Kirk's death was rooted in the American context, it became a part of a broader, transnational discourse. The clustering of engagement within Anglophone countries also highlights the linguistic and ideological pathways through which American political discourse can travel internationally and how global audiences selectively participate in it.

### **2. How did reactions to the event vary between countries?**

To understand not just where the discussion originated, but how it unfolded, we analyzed how content creators positioned themselves with respect to key themes associated with the incident. We identified three topics that were both highly relevant and frequently mentioned across countries, with an overview of the nature of the conversation below:

- 1. **Charlie Kirk (6,224 videos)**—The polarized reception of Charlie Kirk and his legacy
- 2. **Donald Trump (667 videos)**—The U.S. political climate under Donald Trump, and Charlie Kirk's connection to Trump and the MAGA movement
- 3. **Political Violence in the U.S. (3,881 videos)**—The perceived state, (il)legitimacy, and implications of political violence in the U.S.

For each video, we assessed if one or more topics were present. Then, for each topic, we classified the stance expressed toward the topic as in **favour**, **against**, or **neutral**. A 'favour' stance indicates that the video expressed sympathy, approval or support for the topic (e.g., praising Kirk or endorsing Trump). An 'against' stance captures disagreement, criticism, or condemnation (e.g., denouncing Kirk's rhetoric or rejecting violence). 'Neutral' indicates that the poster mentioned the topic without using overtly positive or negative language (e.g., introducing news about the assassination, the suspect, or the memorial without commenting). Importantly, this does not necessarily mean they hold neutral views on the topic. Stances were automatically classified using validated LLMs.

For each topic, we compared the share of videos expressing positive, negative, and neutral stances across the five countries with the highest volume of discussion, grouping all other countries into an 'Other' category. This strategy helps avoid misleading interpretations that could arise from countries with few videos on a given topic. As illustrated in [Figure 3](#page-42-0), these topics capture three dominant dimensions of the online discussion related to Charlie's assassination:

**Charlie Kirk:** Audiences were divided in their views of Kirk, what he stood for, and his legacy. Overall, posts were roughly evenly split among positive, negative, and neutral stances.

**Donald Trump:** Discussion of Donald Trump showed a similar polarized structure. Stances were generally more positive (favor) in the U.S., Brazil, and Mexico, than in Spain or the UK, where mentions of Trump were predominantly negative. Mentions of Trump were comparatively infrequent in our subset of videos originating from Canada.

**Political Violence in the U.S.:** In contrast, attitudes toward political violence were consistent across the board. Across all countries, users overwhelmingly expressed negative stances (against) toward violent political actions.

Overall, this analysis highlights how the assassination of Charlie Kirk has generated moral and emotional reactions globally, not only around Charlie Kirk himself, but also high-level political figures like U.S. President Trump, and political violence in democratic politics. The widespread condemnation of violence on TikTok, despite ideological division on other topics, underscores how U.S. political developments can serve as catalysts for global reflection on the use of political violence in politics and the fragility of democratic norms.

### **Conclusion**

The international discussion and engagement with Charlie Kirk's assassination on TikTok demonstrates how far U.S. political events now resonate beyond American borders. Our findings show that the global sum of users produced nearly as many videos – and drew comparable engagement – as American users, highlighting how major U.S. political shocks can quickly become a shared topic of interest across the broader online information ecosystem. This was particularly visible in English-speaking countries such as Canada, Australia, and the U.K., where shared language, media environments, and cultural proximity appear to facilitate interest in and engagement with U.S. political events.

Overall, online reactions to the event reflected both polarization and convergence. On one hand, opinions toward Charlie Kirk and Donald Trump were largely divided, underscoring the polarized nature of both figures as public personas, and how partisan identities can travel internationally through networks of media and influencers. On the other hand, global online discourse overwhelmingly condemned political violence. This suggests that while political polarization exists globally, it does not extend to attitudes toward political violence, and norms against violence remain widely upheld.

Taken together, this analysis shows that U.S political shocks can trigger large-scale, cross-border conversations, especially within Anglophone spaces, and provoke moments of international moral reflections.

### **Appendix**

### **Prompt used for stance detection:**

**System Prompt:** "You are an expert at analyzing stances in documents.",

**User Message:** """Your task is to determine the stance expressed towards a specific target in the given document. Consider both explicit and implicit indicators of stance.

### Instructions:

- 1. Carefully read the document while focusing on content related to the provided stance target.
- 2. Classify the stance as one of:
- FAVOR: Supporting, promoting, or agreeing with the target
- AGAINST: Opposing, criticizing, or disagreeing with the target
- NEUTRAL: Presenting balanced or objective information about the target

### Input:

Document: [text]

Stance target: [noun phrase]

Output format:

Stance: [FAVOR/AGAINST/NEUTRAL]

 Reasoning: [Brief explanation citing specific evidence from the text]

Examples:

Input:

 Document: "Research shows that diverse communities have higher rates of innovation. Cities with more international residents see increased patent filings and startups."

Stance target: immigration""",

**Prefilled Model Response:** """Output:

Stance: FAVOR

 Reasoning: Text implicitly supports immigration by highlighting its positive economic impacts through innovation and business creation.

""",

**User Message:** """Input:

 Document: "Tech companies want self-governance of social media, while lawmakers push for oversight. Recent polls show the public remains divided."

 Stance target: social media regulation""",

**Prefilled Model Response:** """Output:

Stance: NEUTRAL

 Reasoning: Presents both industry and government perspectives without favoring either side.

""",

**User Message:** """Input:

 Document: "Standardized test scores correlate more with family income than academic ability, while countries using alternative assessments report better outcomes."

Stance target: standardized testing""",

**Prefilled Model Response:** """Output:

Stance: AGAINST

 Reasoning: Implies tests are flawed by linking them to wealth rather than ability and noting superior alternatives.

""",

**User Message:** """Input:

 Document: "Some remote workers report higher productivity, others struggle with collaboration. Companies are testing hybrid models."

Stance target: remote work""",

**Prefilled Model Response:** """Output:

Stance: NEUTRAL

 Reasoning: Balances positive and negative aspects of remote work without taking a position.

""",

**User Message Template:** f"""---

Document: "{doc}"

Stance target: {stance\_target}""",

**Prefilled Model Start:** """Output:

Stance: """

### <span id="page-45-0"></span>**Incident Update 3**

# **Shared anxiety, different lenses: Comparing Canadian and American perceptions of political violence**

**Chris Ross, Mathieu Lavigne, Esli Chan, Diya Jiang Media Ecosystem Observatory**

### **Key takeaways:**

- E **Shared recognition that political violence is serious, but Canadians view it as a much greater problem in the United States:** Most Canadians and Americans view Charlie Kirk's assassination as part of broader societal problems, agree political violence is a serious problem in the U.S., and expect it to worsen. Canadians perceive political violence as far more pressing in the U.S. than at home, both in seriousness (89% vs. 50%) and expected worsening (62% vs. 35%).
- E **Partisan patterns differ across borders in the perception of political violence:** In the U.S., Democrats are the most pessimistic about political violence worsening (86%), while in Canada, the highest concern comes from those who identify as Conservatives (50%).
- E **Young Canadians are more likely to see political violence as a means to achieve social change**: 33% of Canadians aged 18–34 view political violence as sometimes necessary for social change, potentially reflecting growing frustration with conventional political channels of representation.
- E **Concern about political violence deters democratic participation**: Roughly half of Canadians and Americans surveyed report they will avoid at least one form of political participation, such as putting up political signs, posting opinions online, or attending protests, because of concern about political violence.

### **Introduction**

### **Context**

Charlie Kirk's assassination, an act of political violence, intensified national debate about polarization and free speech. Following his assassination, there has been a marked increase in public discourse around perceptions of threat, justification, and consequences of political violence in Canada and the U.S. In this incident update, we examine public attitudes towards political violence, assessing whether the incident and ensuing debates are perceived by Canadians as a shared concern or a uniquely American issue.

### **Key questions**

This report evaluates the following research questions, drawing comparisons with U.S. data where available:

- 1. Is Charlie Kirk's assassination perceived as an isolated incident?
- 2. How concerned are Canadians about political violence?
- 3. To what extent do Canadians believe political violence can be justified?
- 4. Does concern about political violence impact participation in public life?

# **Approach and Considerations**

To understand American attitudes in our comparative analysis, we replicate and extend survey research by the [Polarization Re](https://polarizationresearchlab.org/2025/09/23/report-americans-reject-political-violence-but-see-it-as-an-imminent-threat/)[search Lab](https://polarizationresearchlab.org/2025/09/23/report-americans-reject-political-violence-but-see-it-as-an-imminent-threat/) (to be referred to as PRL onwards) conducted the week after Kirk's shooting in Canada. Their data is from a nationally representative sample of 1000 Americans and is indicated by an asterisk (\*) in any plots throughout this update. In one section, we also cite data from a representative survey of 2101 Canadians conducted by the [Digital Public Square](https://digitalpublicsquare.org/insights/survey-from-the-public-safety-project-comet) in 2024 to show possible time trends in the Canadian context.

Our analysis of Canadian attitudes is based on original survey data from a nationally representative sample of 1,496 Canadians, fielded between September 30th and October 4th, 2025, and largely replicates the PRL question set. To ensure that respondents had a shared understanding of political violence, we began our question block on political violence by providing the following definition:

*"By political violence, we mean acts of physical harm, threats of physical harm, or damage to property that are carried out on purpose to influence politics or achieve political goals."*

<span id="page-46-0"></span>**Figure 1. Comparing Canadian and American views on whether Charlie Kirk's assassination reflects a broader societal problem**

<span id="page-46-1"></span>**Figure 2. Comparing Canadian and American perceptions of the threats posed by political violence**

The following analysis should be read with this definition in mind.

### **1. Is Charlie Kirk's assassination perceived as an isolated incident?**

Online conversations about Charlie Kirk's assassination have been wide-ranging. One facet centered on whether the incident was an isolated event or whether it was part of a longer history of politically—motivated acts of violence, in a context of rising polarization and declining trust in democratic institutions.

To assess whether Canadians share Americans' assessment of the incident, we replicated PRL's survey question, asking whether the shooting of Charlie Kirk represents a broader problem in American society or is an isolated event. [Figure 1](#page-46-0)  shows that more than 3 in 4 Americans found the incident to be representative of a "broader societal problem," a view shared by a similar proportion of Canadians (77%). Only a small minority considered it an isolated event (Canada: 8%; U.S.: 12%) or were unsure (Canada: 15%; U.S.: 13%). Simply stated, Canadians and Americans are aligned in viewing Kirk's death as a part of a larger problem in the U.S. and not a one-off tragedy.

### **2. How concerned are Canadians about political violence?**

When directly asked whether political violence is a serious problem, Canadians' and Americans' assessment of the U.S. is strikingly similar. As [Figure 2](#page-46-1) illustrates, 89% of Canadians think political violence is a 'somewhat' or a 'very big' problem in the U.S., a proportion statistically indistinguishable from the 92% of Americans who view it that way. The notable difference lies in Canadians' assessment of Canada: using the same measure, only half (50%) consider political violence a serious problem in Canada.

Using the same survey question, we next examine partisan differences in Canadians' and Americans' perceptions of political violence in their respective country. [Figure 3](#page-47-0) shows that all respondents grouped by partisanship in the U.S. overwhelmingly believe that political violence is a serious problem (93–97%) while the Independents are the least likely to do so (85%). In Canada, concern is generally lower, but Conservatives are significantly more likely to view political violence as a serious problem (61%) as compared to Liberals, New Democrats, or those with no party affiliation (43–48%). Overall, in the post-Kirk

<span id="page-47-0"></span>**Figure 3. Partisan breakdowns of Canadian and American perceptions of the threats posed by political violence**

<span id="page-47-1"></span>**Figure 4. Comparing Canadian and American assessments of whether political violence will get worse or better in the next few years**

assassination period, left-partisan individuals are less likely than right-partisan counterparts to view political violence as a serious issue, particularly in Canada.

Next, we examine whether Canadians and Americans expect political violence to get better or worse in the coming years. [Figure 4](#page-47-1) shows Americans are the most pessimistic about the future of political violence in their own country, with 76% thinking the problem will worsen. Canadians are slightly less pessimistic about the U.S.'s future, though a majority also expect the situation to get worse (62%). Turning to their own country, 35% think political violence will worsen, while nearly half (46%) expect it to remain about the same. These patterns show that Canadians' expectations differ sharply depending on whether they are considering political violence in the U.S. versus Canada.

[Figure 5](#page-48-0) evaluates partisan differences in expectations about whether the problem of political violence will get better or worse in the coming years. In the U.S., Democrats are more likely than Republicans to believe that the problem will worsen (86% versus 68%), while in Canada, Conservatives are more pessimistic than the Liberals (50% versus 28%). These differences are more pronounced than those observed in [Figure 3](#page-47-0) (general perception of political violence as a problem), suggesting that perceptions of future safety and stability are more tied to partisanship.

### **3. To what extent do Canadians believe political violence can be justified?**

Political violence is often discussed in highly charged contexts, where individuals navigate a tension between democratic norms opposing violence and social media discourse praising certain acts of political violence. For example, in December 2024, [Luigi](https://en.wikipedia.org/wiki/Luigi_Mangione)  [Mangione](https://en.wikipedia.org/wiki/Luigi_Mangione) murdered a U.S. healthcare company CEO. His actions were widely discussed on social media, with much of the content glorifying him as a vigilante.

[Figure 6](#page-48-1) examines changes in Canadians' views on whether political violence is sometimes necessary for social change, comparing responses after the Charlie Kirk incident with those collected by the [Digital Public Square](https://digitalpublicsquare.org/insights/survey-from-the-public-safety-project-comet) (DPS) in the summer of

<span id="page-48-0"></span>**Figure 5. Partisan breakdowns of Canadian and American assessments of whether political violence will get worse or better in the next few years**

<span id="page-48-1"></span>**Figure 6. Canadian perceptions of whether political violence is sometimes necessary for social change over time and by age group**

2024. The proportion expressing this belief increased slightly from 16% in 2024 to 21% in 2025, but remains close to previous levels ([20% in 2023](https://digitalpublicsquare.org/insights/survey-from-the-public-safety-project-comet)), indicating overall stability in Canadians' rejection of political violence.

Our data show significant age differences, slightly larger than those identified by the Digital Public Square. Between 2024 and 2025, the share of Canadians aged 18–34 who consider political violence sometimes necessary rose from 26% to 33%, compared with smaller increases among Canadians aged 35–54 (17% to 21%) and those 55 and older (11% to 12%). It is important to note that respondents may have had different forms of violence in mind — ranging from property damage to physical harm — which carry very different implications. That said, these findings invite reflection on how the legitimacy and effectiveness of democratic processes can be strengthened, and on whether those who see political violence as necessary for social change would instead rely on democratic levers if they trusted their efficacy. An underexplored area of this analysis is whether Canadians believe these levers provide real opportunities for change or view political violence as a 'first-choice' mechanism.

### **4. Does concern about political violence impact participation in public life?**

Political violence can have wider societal effects by discouraging people from participating in public life. [Figure 7](#page-49-0) compares how often Canadians and Americans (the latter measured by the

<span id="page-49-0"></span>**Figure 7. Comparing Canadian and American selfreported likelihood of avoiding political activities due to concerns about political violence**

Polarization Research Lab) report avoiding certain activities due to concerns about political violence.

The most notable finding is that about half of respondents, in both Canada and the U.S., report not changing any of their listed behaviors because of political violence. However, that leaves half the population who express concern about political violence resulting in decreased political engagement. Combined with findings from [Figures 2](#page-46-1) and [4](#page-47-1), this highlights a tension between perception and behavior: Canadians see political violence as a greater problem in the U.S. than Canada, but report avoidance of public participation at rates similar to Americans. This underscores that despite differing perceptions of political violence, Canadians and Americans face a similar climate of caution and self-censorship that limits open civic participation.

In Canada and the US, about one quarter of the population express concern in engaging in acts of political expression due to political violence. These actions include putting up political signs (28 vs. 31%), posting political opinions on social media (28%-30%) or attending a protest (23%-27%), with no meaningful difference between Canadian and American attitudes. Though a small figure, 6% of Canadians report that they will avoid voting in an election due to concerns about political violence. Together, these findings expose a serious strain on democratic participation.

Political violence can thus cause a chilling effect, both online and offline, discouraging people from engaging in visible political action. One additional action we examined—not included in the PRL survey—is the willingness to become a public figure. While few Canadians actually face this choice, citizens' willingness to take on a public role is a vital dimension of a healthy, participatory democracy. Alarmingly, about one in five

Canadians (21%) report that concerns about political violence make them likely to avoid becoming a public figure. When fear of political violence dissuades Canadians from pursuing political office or engaging publicly, it weakens the inclusivity of voices in democratic discourse and erodes the resilience of democratic institutions.

### **Conclusion**

Our findings suggest that Canadians and Americans share a deep concern about the state of political violence and its broader social implications. The vast majority in both countries see the Charlie Kirk assassination as a reflection of wider societal problems. 9 out of 10 Canadians and Americans agree that political violence is a major issue in the U.S. and that the situation is likely to get worse in the coming years across the board.

At the same time, our results reveal significant partisan divides across the border. In both countries, people from the opposition party (Democrats in the U.S. and Conservatives in Canada) express greater pessimism about the future of political violence, suggesting that such views can be tied to trust in the current leadership. Although Canadians across age groups generally reject political violence, younger adults stand out for viewing political violence as potentially necessary for social change. Moreover, citizens in both countries report avoiding democratic expression due to concerns of political violence. Together, these patterns indicate that the shadow of political violence shapes how people perceive the state of their democracy, dissent, and political participation.

### <span id="page-50-0"></span>**Incident Update 2**

# **Online Shockwaves: Reactions to the Charlie Kirk Incident on X in Canada and Abroad**

**Zeynep Pehlivan, Esli Chan Media Ecosystem Observatory** 

### **Key takeaways**

- E **Lifespan of conspiratorial and partisan narratives was short**: During the initial shock of the shooting, conspiratorial and partisan narratives emerged online globally, but quickly faded in engagement. This demonstrates how outrage-based narratives can emerge and fade just as quickly.
- E **Clear topic segregation between news outlets, influencers, and politicians in Canada:** During the peak of the Kirk-shooting discourse, news outlets provided key informational updates while influencers guided the political commentary, and politicians picked and chose their spots engaging when it fit official narratives and leaving conten-

tious or ritualized threads to newsrooms and creators. This reflects a division of communication roles, with each group contributing differently to the overall flow and interpretation of information.

E **While Canadian conversation mirrored the US, attention was limited:** The thematic structure of discourse in Canada closely mirrored trends in the US; yet, the volume and degree of engagement was comparatively reduced. This demonstrates that while the Canadian information ecosystem is highly susceptible to political events in the U.S., its distinct media structures and actor dynamics can help moderate the intensity and persistence of heavily polarized online debates.

# **Introduction**

In this incident update, we evaluate how [the death of Charlie Kirk](https://www.cdmrn.ca/publications/effects-of-the-charlie-kirk-assassination-in-the-canadian-information-ecosystem), a prominent right-wing American commentator, has provoked widespread online debate. Controversy around the legacy of Kirk has sparked discussion around the increasing partisan divide, ideological polarization, and threats of political violence. This discourse, while originating from an American context, has spilled over into the Canadian information landscape and subsequently influenced the health of the online space (e.g., insularity, segmentation, polarization). Our evaluation focuses on the topics that were most prominent and how they evolved into different interpretive narratives over time. We evaluate who captured most of the attention and shaped the conversation; and how partisan framing, free-speech arguments, and foreign-influence claims appeared in the discussion. Further, we assess

<span id="page-51-0"></span>**Figure 2. Volume of posts mentioning "Kirk" on X from September 5–21, 2025, segmented into four phases: Phase 1—Initial Framing & Shock;** 

**Phase 2—Partisan Polarization; Phase 3— External Narrative Dominance; Phase 4— Sustained Narrative Contestation.** 

how discourse around the Kirk shooting spread throughout the Canadian information ecosystem (CIE).

To do this, we created a dataset with the posts on X (formerly Twitter) mentioning "kirk" from September 5 to 21, 2025. Figure 1 illustrates the distribution of posts over time, with over 100,000 posts on the day of shooting on September 10, 2025.

# **Key questions**

Our key questions focus on understanding the key topics and narratives that shaped the online discourse around the Charlie Kirk assassination. We ask the following two questions:

- 1. What did people talk about globally?
- 2. What did Canadians talk about?

### **1. What Did People Talk About Globally?**

We wanted to know which narratives dominated around the world, and how they rose and faded over time? To answer these questions, we assessed how different topics of conversation arose on X throughout the days following Kirk's death, and the types of narratives used to frame and portray the event. To identify and group posts discussing similar themes, we employed topic modeling, a computational method that analyzes a large collection of text to automatically group together and identify recurring themes or subjects discussed in the ecosystem.

We find that the conversation on X quickly became a polarized battleground, shifting from immediate conspiracy theories and

grief to sustained ideological warfare. Contrasting the number of posts (post volume) to engagement with these posts (likes), the dominant topics varied between the two. While high post volume centered on Israeli government involvement, the Epstein files, and the shooter's identity, audience reactions (likes) favored different topics. References to Elon Musk, Christianity/faith, and the Israeli government's response received the highest overall engagement and were exceptionally efficient at generating per-post likes, making them clear narrative outliers. The political Left's response, the (temporary) cancellation of the Jimmy Kimmel show, and Erika Kirk's public statements also generated significant attention, primarily driven by right-wing political commentators and news outlets**.**

### **The Four Phases of Narrative Evolution**

We identified four distinct overlapping phases of public discourse on X following Kirk's assassination. These phases include:

**Phase 1: Initial Framing & Shock**—High engagement framed around conspiratorial and outrage-based narratives.

**Phase 2: Partisan Polarization**—Left- versus right-leaning critique on responses to Kirk's death.

**Phase 3: External Narrative Dominance**—Side conversations related to Kirk that emerged about wider cultural touchpoints and fallout.

**Phase 4: Sustained Narrative Contestation** — Persistent engagement on freedom of expression as a result of criticism versus support of Kirk's legacy.

<span id="page-52-0"></span>**Table 1. The four phases.**

| Phase                                                                      | Phase 1:<br>Initial Framing & Shock                                                                                                                                                                                                                                                                                                                                                   | Phase 2:<br>Partisan Polarization                                                                                                                                                                                                                              | Phase 3:<br>External Narrative<br>Dominance                                                                                                                                                                                                                                                                                                                                                                           | Phase 4:<br>Sustained Narrative<br>Contestation                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                         |  |
|----------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|--|
| Dominant Topics of<br>Discussion                                           | E Transgender identity of<br>the shooter                                                                                                                                                                                                                                                                                                                                              | E Liberals and Democrats'<br>responses                                                                                                                                                                                                                         | E The cancellation of<br>Jimmy Kimmel's show                                                                                                                                                                                                                                                                                                                                                                          | E Debates around free<br>speech                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                         |  |
|                                                                            | E Israeli Government's in<br>volvement in Kirk's death                                                                                                                                                                                                                                                                                                                                | E Celebration of Kirk's<br>death                                                                                                                                                                                                                               | E Erika Kirk's response                                                                                                                                                                                                                                                                                                                                                                                               | E Critiques of Kirk's polit<br>ical stances (e.g. DEI)                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                  |  |
|                                                                            | E Relation to the Epstein<br>Files                                                                                                                                                                                                                                                                                                                                                    |                                                                                                                                                                                                                                                                |                                                                                                                                                                                                                                                                                                                                                                                                                       | E Empathy and support<br>for Kirk's family                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                              |  |
| Nature of Discussion                                                       | Emerged instantaneously<br>from the shock, focusing<br>on the killer's link to their<br>potential transgender<br>identity and intent. The<br>topics leveraged the initial<br>shock for conspirator<br>ial and outrage-based<br>narratives. Attention was<br>driven by Netanyahu's<br>quick engagement and<br>claims that the shooting<br>was a distraction from the<br>Epstein files. | A partisan lens to frame<br>the political Left and<br>Democrats as respon<br>sible for Kirk's death and<br>societal violence. Political<br>officials and commenta<br>tors acted as critics, fram<br>ing the celebration of Kirk's<br>death as morally corrupt. | Discussion shifted beyond<br>the original event towards<br>subsequent cultural touch<br>points triggered by the<br>death. Topics included the<br>suspension of Jimmy Kim<br>mel's show and Erika Kirk's<br>public presence as the new<br>CEO of Turning Point USA<br>(TPUSA) and her forgive<br>ness of the shooter.                                                                                                  | This phase is characterized<br>by a steady and under<br>lying stream of discourse<br>for the following topics,<br>demonstrating ongoing<br>debate. The Right (ampli<br>fied by figures like Trump)<br>framed Kirk as a martyr<br>of free speech, while the<br>Left highlighted perceived<br>hypocrisy in right-lean<br>ing free speech advocacy<br>given their crackdown on<br>DEI-related educational<br>materials. Free speech<br>became a persistent<br>battleground used by<br>both sides to contest the<br>other's credibility. Sup<br>port for Kirk's family was<br>largely used empathetic<br>language, e.g., "praying for<br>the Kirk family" . |  |
| Volume (number of<br>posts) and Engage<br>ment (average likes<br>per post) | Posts surged on September<br>10 (the day of the shooting)<br>to approximately 8,000<br>posts and spiked at around<br>10,000 posts on Sept. 11 (the<br>day after the shooting).<br>Posts dropped sharply one<br>week later (Sept. 18) Con<br>spiracies attracted quick,<br>short-lived attention.<br>Engagement was high<br>early on but dropped<br>sharply as speculation<br>faded.   |                                                                                                                                                                                                                                                                | Two delayed spikes re<br>vealed a revival in the nar<br>also at roughly 8,000 posts,<br>rative, with posts surging<br>on September 17–18 (Kim<br>mel's suspension/Erika<br>Kirk's public statement)<br>and again on September<br>21 (Erika Kirk's message<br>of forgiveness during the<br>memorial).<br>Despite limited posts, en<br>gagement peaked on Sep<br>tember 12 with Erika Kirk's<br>first public statement. | Posts emerged with Phases<br>1 and 2, with slightly less<br>volume (over 5,000 posts)<br>and held a stable average<br>of posts per day until Sept.<br>20.<br>Engagement fluctuated<br>over time but maintained<br>at a moderate level of<br>1,000–4,000 likes per day.                                                                                                                                                                                                                                                                                                                                                                                  |  |

<span id="page-53-0"></span>**Figure 3. Likes (Engagement) on posts including "Kirk" on X from September 5 to 21nd, 2025 Phase 1— Initial Framing & Shock; Phase 2—Partisan** 

**Polarization; Phase 3—External Narrative Dominance; Phase 4—Sustained Narrative Contestation.** 

[Table 1](#page-52-0) below describes each of these phases in more detail (see Methodology section for information about how these phases were identified). For each phase, we describe the dominant topics, the nature of discussion and volume and engagement with the topics. The volume of posts and engagement over time corresponding to each phase are detailed in [Figures 2](#page-51-0) and [3](#page-53-0). The phase determination process is outlined in the methodology section at the end of this update.

### **2. What did Canadians talk about?**

In Canada, what narratives emerged within the Canadian information ecosystem? Which actors led the conversation? How did discussion differ from the global dynamics? To answer, we analyzed Canada's information ecosystem from Sept. 5–21. We identified 5,308 X posts mentioning "Kirk" from our existing Canadian seedlist, amounting to about 4% of total CIE activity.

**When did the conversation peak in Canada?** Examining the daily volume of posts, Figure 4 shows an early condolences spike (Sept. 10), a secondary peak related to government criticism and/ or celebrations (Sept. 11), and, as offline gatherings reignited attention, a third peak, associated with vigils. As compared to the global conversational trends, Canadian engagement follows the same initial interest in the topic with increased engagement on September 10 and 11. However, Canadian online activity decreases at a relatively higher rate — while global engagement is prolonged5.

**Who talked about what?** The most prominent topics in Canada following the event include: the Canadian government's response, vigils and memorials happening across Canada, condolences for Kirk and his family, and updates and reactions to Trump's response to the event. However, the emphasis on these topics varied significantly across actors.

[Figure 5](#page-54-0) outlines the most prominent topics influencers, news outlets, politicians and civil society organizations posted about (posts) and engaged with (likes) over the period of study. Examining the left panel of Figure 5, *News outlets* acted as the main information source, providing informational updates regarding the arrest of the suspect and evolution of the incident. *Influencers* were primarily commentators, dominating the conversation related to the celebration and condemnation of Kirk's legacy and how the Canadian government responded to the event. *Politicians* were noticeably absent in discussion around polarization and blame, vigils and memorials happening in Canada, and the nature of the suspect and their arrest. The lack of engagement from politicians left space for newsrooms and creators. Lastly, civil society organizations engaged the least those who posted were largely right-leaning Canadian groups who remarked on the positive legacy of Kirk and condemned political violence.

**How is engagement distributed?** Looking at the right panel of [Figure 5](#page-54-0), *Politicians* captured the most attention on topics such as the alleged involvement of the Israeli government (98.7% of the topic share), media updates (78.0%), and condemnation of

**Figure 4. Distribution of number of posts for top 5 topics in the Canadian Information Ecosystem**

<span id="page-54-0"></span>**Figure 5. Account-normalized posts and likes by topics6 in the Canadian Information Ecosystem**

political violence (47.8%), setting the groundwork for high-salience political debates. *News outlets* then amplified and framed these issues, gaining attention on updates regarding Kirk's previous statements related to identity politics (67.3%) and Erika Kirk's public statements (70.6%), thereby legitimizing certain narratives and shaping how issues are interpreted. *Influencers*, in turn, drove reactions within commentary spaces by translating these political and media narratives into affective, emotionally resonant stories that circulated within their online communities.

These findings demonstrate how different actors in the online ecosystem each play unique roles in diffusing and amplifying various elements of a story: news outlets often establish the factual foundation, influencers amplify emotional and interpretive angles, and political actors largely contribute to meaning-making. Together, they create a multi-layered and emotionally charged information environment, where facts, reactions, and meaning-making circulate and reinforce one another.

### **Conclusion**

The online discourse surrounding the death of Charlie Kirk on X demonstrates the dynamic and polarized nature of contemporary political narrative diffusion. The incident did not result in a single, unified conversation but rather a four-phased evolution characterized by distinct patterns of volume and engagement.

In Canada, the conversation was small, only about 4% of the total X activity analyzed, but mirrored the polarized structure. The Canadian information ecosystem's response to Kirk's shooting showed a sophisticated and layered form of polarization: news accounts moved facts quickly while influencers/politicians pushed partisan frames. The Canadian space is shaped by U.S. discourse, but its impact is moderated by local structures and a smaller, less-engaged politician cohort, which limits the intensity compared to the U.S.

# **Methodology**

### **Topic modeling**

The present study adopts BERTopic, a modular pipeline that combines transformer-based embeddings with non-parametric clustering to yield coherent, human-interpretable topics. In the first stage, each text instance is converted into a high-dimensional embedding using the multilingual variant of MiniLM fine-tuned for paraphrase similarity. These embeddings are then projected to a low-dimensional manifold with UMAP, which preserves the local neighbourhood structure while substan-

tially reducing computational complexity for subsequent steps. Topic discovery is performed with HDBSCAN, a density-based algorithm that automatically determines the number of clusters and isolates noise points; its minimum cluster size parameter is scaled to corpus size to balance granularity and stability.

Once clusters are identified, BERTopic constructs a sparse class-based TF-IDF (c-TF-IDF) representation that contrasts the vocabulary of each cluster with the entire corpus, thereby highlighting discriminative terms even in highly imbalanced data. To enhance interpretability, the framework ensembles multiple representation strategies: a KeyBERT-inspired extraction ranks candidate keywords by their cosine similarity to the centroid embedding, while a Maximal Marginal Relevance step re-orders them to maximise semantic coverage and minimise redundancy. The top keywords form concise, descriptive labels that are appended to the numerical topic identifier, and each document inherits the label of its most probable cluster together with an assignment confidence.

### **The phase determination process**

The 11 most active topics, collectively covering 90% of the data, were analytically regrouped into four distinct Narrative Phases to assess the evolution and competition between different thematic arcs.

The phase determination process moves beyond simple chronological blocks and is grounded in two core criteria:

**Temporal Co-dominance (Quantitative):** We identified time periods where specific clusters of topics exhibited a synchronous spike in post volume, indicating a shared, temporary shift in collective attention. This was used to establish the preliminary boundaries where the topic distribution changed drastically.

**Narrative Cohesion (Qualitative):** Topics identified in the same time frame were grouped only if they shared a unified thematic purpose or were directly related by cause-and-effect in the ensuing debate.

This approach treats the "phases" as analytical constructs designed to group competing narrative arcs that inform and react to one another, offering insight into the overall structure of the discourse rather than a strict, linear timeline.

### <span id="page-56-0"></span>**Incident Update 1**

# **Kremlin Doppelgänger Tactics and the Exploitation of the Charlie Kirk Murder**

**Marcus Kolga, DisinfoWatch Layla Mashkoor, DFRLab Meredith Furbish, DFRLab**

### **Key takeaways**

- E **Strong evidence of information warfare tactics employed by Russia:** With high confidence, it appears that though the Kremlin has exploited Kirk's death as part of its information warfare strategy, employing tactics outlined in the FBI's Doppelgänger affidavit.
- E **Amplified claims were mostly conspiratorial in nature:** Dominant claims included making links to historic U.S. assassinations, emphasizing that the American left celebrated his death, and links tying the killing to Ukraine or Israel.
- E **Russian response aimed to polarize and delegitimize:** Information operations involved monitoring foreign information spaces to identify emerging issues most likely to fracture society, and then deploying a suite of tactics to circulate and amplify false or contradictory narratives and conspiracies, with the aim to deepen divisions and polarization.

**Figure 1. Screenshot of the FBI Doppelganger Affidavit** 

### **Introduction Context**

In September 2024, the FBI released an [affidavit](https://www.justice.gov/archives/opa/pr/justice-department-disrupts-covert-russian-government-sponsored-foreign-malign-influence) detailing Russia's Doppelgänger campaign, an information operation that deployed networks of counterfeit URLs mimicking major Western media outlets in order to deceive social media users worldwide. The affidavit included classified documents from the Russian Presidential Administration that outlined the Kremlin's strategic objectives and information warfare and influence operations tactics, techniques and procedures. These materials, consisting of strategic plans and minutes from meetings chaired by Vladimir Putin's senior advisors, directed three Russian communications firms, Social Design Agency (SDA), Structura National Technology (Structura), and ANO Dialog, to conduct influence operations abroad.

The documents instructed these firms to closely monitor Western information and political environments and to identify events and issues with the greatest potential to divide and polarize democratic societies. These efforts formed part of the Kremlin's "International Conflict Incitement" project, whose stated goal was to "influence real-life conflicts and artificially create conflict situations." Operational tactics involve saturating the information environment with narratives designed to confuse and divide audiences through methods such as "traffic redirection, coordinated comment campaigns, engagement with influencers, analytical articles, augmented reality, media mirror outlets, fake content, and more," all with the ultimate objective of undermining societal cohesion. Contractors and agents were further directed to disseminate "false narratives," "fabricated videos, documents, and recorded conversations," as well as "social media commentary" and "quotes, both real and invented, from public figures" – in essence, to **saturate**, **confuse** and **exploit** the event.

**Figure 2. Examples of screenshots demonstrating Russian attempts to polarize and legitimize the killing**

These materials illustrate in remarkable detail the Kremlin's systematic approach to operationalizing information warfare. While the Russian Foreign Ministry [explicitly denies trying to](http://web.archive.org/web/20250919193901/https://www.rt.com/russia/624932-moscow-responds-us-media-claims-kirk-murder/)  [exploit the murder](http://web.archive.org/web/20250919193901/https://www.rt.com/russia/624932-moscow-responds-us-media-claims-kirk-murder/), the strategies and tactics identified in the FBI investigation were clearly reflected in the rapid exploitation of Charlie Kirk's murder in September 2025.

### **Key questions**

To assess the nature of Russia's involvement following Kirk's death, we seek to address three questions:

- 1. Did Russia apply information warfare tactics in the wake of Charlie Kirk's assassination?
- 2. What claims emerged from Russian state-sponsored media following the event?
- 3. What is Russia trying to achieve?

### **1. Did Russia apply information warfare tactics in the wake of Charlie Kirk's assassination?**

Following Kirk's death, Russian state media and aligned channels mobilized almost immediately. Explained in more detail in question 2, state outlets seeded politicized and conspiratorial frames that ranged from [blaming](https://archive.is/myfhQ) [Ukraine](https://archive.is/WaYV9) to drawing on [U.S.](https://archive.is/LrDKR)  [culture-war](https://archive.is/LrDKR) [narratives](https://archive.is/M2zvV). This reflects the operational features and tactics described in the FBI's [Doppelgänger](https://mpf.se/psychological-defence-agency/publications/archive/2025-05-15-beyond-operation-doppelganger-a-capability-assessment-of-the-social-design-agency) [affidavit](https://www.justice.gov/archives/opa/pr/justice-department-disrupts-covert-russian-government-sponsored-foreign-malign-influence), which focuses on identifying and exploiting issues with the greatest potential to divide Western societies.

**Rapid saturation.** A flood of stories and posts ensured that Russian narratives entered circulation before facts were established.

**Plural and inconsistent storylines.** Some narratives connected the murder to U.S. [support](https://web.archive.org/web/20250911144114/https://www.rt.com/russia/624511-ukraine-kirk-death-reaction/) for [Ukraine](https://archive.is/qE7ey). Others invoked identity politics or [anti-Zionist](https://archive.is/hl1b1) tropes. Consistency was secondary to the effect of fragmentation.

**Amplification.** Russian state-controlled channels were monitoring the US information space and were selectively amplifying content with the greatest potential to intensify polarization.

The pattern mirrored the doctrine [described in the FBI affidavit](https://www.justice.gov/archives/opa/pr/justice-department-disrupts-covert-russian-government-sponsored-foreign-malign-influence): to **saturate** the information environment, **confuse** audiences, and **exploit** tragedy to deepen social divisions.

### **2. What claims emerged from Russian state-sponsored media following Kirk's death?**

In-depth investigation of Russian state-media outlets, Sputnik and Russia Today (RT), revealed a strong push of conspiratorial claims. While Sputnik devoted only a small amount of coverage to Kirk's assassination, they leaned heavily on narratives particularly linking the event to other famous American assassinations (John F. and Robert F. Kennedy Sr., Martin Luther King, and Malcolm X). [This story](https://archive.is/KB3Rn) was based on an interview with a former FSB agent, who promoted common conspiracy tropes like sleeper agents and the supposed deep state controlling US security forces.

In contrast, RT's X account seized the incident almost immediately, and the focus was disproportionately high in comparison to other subjects. The main claim in their coverage was that the American political left was celebrating Kirk's death (see [here](https://x.com/RT_com/status/1966918555137647081) and [here](https://x.com/RT_com/status/1967248708833358206)). While the assassination was widely condemned across the American political spectrum, over a quarter of RT posts about Kirk in the 24 hours after his death were about unsympathetic or celebratory reactions to his death. Just under 20 percent of posts were about sympathetic responses or grief, but these were almost exclusively the reactions of right-wing figures, which appears to be an attempt to create the impression of a partisan split in attitudes to Kirk's death.

More broadly, RT and other news sources suggested the assassination was connected to Ukraine and, also, Israel. Discussions about Ukraine, aimed at fabricating a connection between the

assailant and [Ukraine](https://archive.is/qE7ey) or those [supportive](https://archive.is/qqvN7) of Ukraine. These narratives included posts that emphasized Kirk's previous [remarks](https://archive.is/Z3fsq) regarding Ukraine, presenting them as a possible motive. Former Russian President Dmitry Medvedev [explicitly](https://archive.is/bWW35)  [accused Ukrainians](https://archive.is/bWW35) in a tweet posted the day of the murder. Concurrently, other content sought to exploit cultural fault lines by advancing claims of a conspiracy involving a so-called "[transgender terrorist cell](https://archive.is/M2zvV)." Discussions about Israel were more subtle. Posts implicated Israeli involvement through imagery like the Star of David or Israeli flags rather than explicit mention, prompting audiences to respond with criticism of Israel and antisemitism, including conspiratorial claims linking Israel to historical assassinations.

### **3. What is Russia trying to achieve?**

The tactics used serve broader strategic objectives. Russian information operations are part of a coherent strategy aimed at achieving three interrelated goals.

**Polarization of Western societies.** Moscow benefits when democratic states are internally divided. By introducing conspiracy, grievance, and culture-war framing into the Kirk case, Russian media sharpened political and social cleavages in the United States.

**Delegitimization of democratic institutions.** Narratives that suggested authorities were concealing information or that media coverage was unreliable aligned with Russia's long-standing goal of undermining confidence in democratic governance and independent journalism.

**Erosion of support for Ukraine.** By associating Kirk's murder with U.S. policy toward Ukraine, and speculating about Kyiv's involvement, Russia advanced its strategic priority of weakening Western resolve to sustain military and political support for Ukraine.

RT's coverage illustrates these objectives. It framed the killing as an "assault on free speech," recycled Kremlin-friendly claims about Ukraine, and amplified identity-based grievances. This was not conventional reporting but a deliberate attempt to instrumentalize the event for political effect.

### **Recommendations**

The Kremlin's exploitation of the Kirk case followed the same operating logic established in its Doppelgänger program: industrial-scale disinformation measured by reach rather than accuracy. By pushing contradictory narratives and ensuring their rapid penetration into public debate, Russia converted a domestic American tragedy into a tool for advancing its foreign policy aims. Canada, like other democracies, was not immune to these narratives.

The documents exposed in the FBI affidavit provide clear insight into Russia's objectives and methods. Governments, media, and civil society should have anticipated that Moscow would seize on the Kirk incident, just as it will exploit future crises and sensitive issues. The Doppelgänger materials should be treated as a roadmap for building defences that reinforce democratic information environments and societal resilience against the manipulation of tragedy.

Recognizing this doctrine and preparing accordingly is essential if democracies are to blunt the impact of future Russian information manipulation campaigns.

### <span id="page-60-0"></span>**Endnotes**

- **1 Lally, Nick. "Crowdsourced Surveillance and Networked Data." Security Dialogue 48, no. 1 (2017): 63–77.**https://www.jstor.org/stable/26294202**. And Favarel-Garrigues, Gilles, Samuel Tanner, and Daniel Trottier. 2020. "Introducing Digital Vigilantism." Global Crime 21 (3–4): 189–95. doi:10.1080/17440572.2020.1750789.**
- **2 "State-affiliated media" disclosures on platforms such as** TikTok**, X and other social media refer to labels assigned to media accounts or content that are financially or editorially linked to governments. For example, on X, the label was used from 2020 to 2023 to flag outlets like Russia's RT or China's Xinhua that are controlled by their governments (the label was** removed in April 2023**). These disclosures are designed to give users transparency about potential state influence on media — helping readers assess context, possible bias or agenda behind content.**
- **3 There is a minor difference in our measure compared to the linked so, while Canadians encountered polarized discourse online, it did not translate into meaningful changes in how they generally view one another. We measure attitudes towards the actual political party whereas Merkley uses the common method of measuring attitudes towards members of said party. These values are generally highly correlated.**
- **4 By claims about user celebration we are referring to "People celebrated the fact that Charlie Kirk died," "People should get fired for celebrating Charlie Kirk's death," "Left-wing individuals celebrated the fact that Charlie Kirk died."**
- **5 We note that the data sources for analysis of the global response (all X posts mentioning 'Kirk') are different from the Canadian response (all X posts mentioning 'Kirk' within our existing Canadian seedlist). The analysis here is a narrative assessment rather than an empirical data comparison.**
- **6 Values show the breakdown by actor type (influencer, politician, etc.) within each topic. For each topic × actor type, we count posts per account and likes per account, take the mean for that actor type, then rescale within the topic so shares sum to 100%. (Applied separately for posts and for likes.)**