Report finds social media ‘amplifying anti-Muslim hatred’ in Ireland after Dublin mosque fire went viral
New study by Hope and Courage Collective found that video of alleged arson at Dublin mosque was viewed 1.9 million times and that one in five of comments contained incitement or calls for further attacks
By MNTV Staff Writer
When a Facebook account posted a video of firefighters at the scene of an alleged arson attack on a Dublin mosque, it did not simply document an incident. It opened a floodgate.
The clip was watched more than 1.9 million times, drew over 10,600 comments, and became a rallying point for hundreds of open calls to burn more mosques and kill Muslims — content that, weeks later, largely remained online.
That is the central finding of a new report by the Hope and Courage Collective (HCC), an Irish civil society organisation that works to counter far-right disinformation and hate speech.
Titled The Islamophobia Playbook: A Project of Hate, the report presents a detailed case study of the online response to the alleged arson at the Al Madinah Prayer Hall on Talbot Street in Dublin, and argues that the design of major social media platforms — not merely the presence of bigoted users — is actively amplifying anti-Muslim hatred in Ireland.
Video that outperformed everything
The mechanics of amplification are laid bare in the report’s numbers. According to the HCC, a Facebook profile called “East Wall Protests” posted a reel on June 29 combining two clips of Dublin Fire Brigade at the mosque, captioned simply that a mosque on Talbot Street was on fire.
The reach was extraordinary — and anomalous. The video drew more than 1.9 million views, which the report calculates at roughly 352 times the account’s normal audience.
The account’s videos typically average around 5,400 views; this one represented what the researchers describe as a 35,000 per cent increase, sitting far outside the expected reach of the profile’s content.
The account, set up in November 2022, had from April 2024 almost exclusively posted material supporting a particular Dublin City councillor.
For the HCC, that gap between a small account’s usual footprint and this single video’s vast reach is not incidental — it is the story. It points, the researchers argue, to the role of platform recommender systems in deciding which content is pushed into users’ feeds, and to a system that rewarded a video attached to an anti-Muslim attack with exponential visibility.
One in five comments: hate, incitement, and worse
If the reach was alarming, the report found the response beneath the video more so. Researchers analysed more than 10,600 comments and emoji reactions and concluded that one in five related to violence and incitement — including calls to target specific locations, anti-Muslim hate speech, exclusionary and conspiratorial narratives, and religious hatred.
The specifics are grim. Scores of comments called for mass violence against Muslims in Ireland.
More than 300 explicitly called for further mosques to be firebombed or burned down. The phrase “burn them all” recurred throughout. The video itself had attracted over 31,000 likes.
The report documents how the alleged attack was not merely excused but celebrated, with hundreds of comments — “one down,” “more to go,” “let it burn” — that, while not direct calls for further violence, had the effect of normalising arson against a place of worship.
Alongside these ran explicit calls to deport Muslims en masse, to close mosques, to ban the practice of Islam, and a steady stream of dehumanising language likening Muslims to “rats” and “cockroaches” that needed to be “burned out” — rhetoric with a dark historical resonance that the report does not shy from noting.
Coded language of hate
One of the report’s more analytically striking sections concerns how hatred is disguised.
Researchers found widespread use of what they term coded humour — recurring references to bacon, pork, rashers, pigs, and barbecues — which, far from being harmless jokes, functioned as a shared language mocking Islamic religious practice while celebrating the burning of a mosque.
Emojis played a similar role. Of the 10,600 comments examined, more than 3,600 used emojis; over 1,540 used laughter emojis and more than 560 used fire emojis.
In the context of a video about a mosque fire, the report argues, these symbols allowed support for anti-Muslim violence to be expressed indirectly — a finding consistent with earlier HCC research into arson attacks.
The effect is a form of plausible deniability, hatred rendered as banter, that makes the content harder to moderate and easier to spread.
Platforms in frame: Meta and X
The report reserves its sharpest questions for the platforms themselves.
On Meta, the HCC argues the evidence points to a systemic failure to enforce the company’s own rules — specifically at least three separate policies covering coordinating harm and promoting crime, hateful conduct, and violence and incitement.
This was not, the report stresses, a matter of a few comments slipping through: the scale demonstrates systemic failures and risks arising directly from how Meta operates. The organisation is calling on Ireland’s media regulator, Coimisiún na Meán, to press Meta to publish an incident-specific risk assessment explaining how it responded to the Talbot Street incident.
Mark Malone, Research Lead at the Hope and Courage Collective, said Meta needed to answer directly why it continued to host calls for further attacks on mosques and for violence against Muslims, and demanded that the company “demonstrate transparently how their systems are assessing and mitigating systemic risks created by their own recommender systems.”
He was scathing about the company’s standard denials: “Copy and paste statements from Meta about how it does not allow hate speech on its platforms do nothing to address the reality that it does.”
In a statement to RTÉ, a Meta spokesperson said the company does not allow hate speech on its platforms and has clear policies prohibiting content that attacks people based on protected characteristics including religion and ethnicity, adding that it takes action on content that violates those policies. The company said its recommendation systems are built with integrity safeguards, that content breaching its community standards is removed, and that some material is permitted to remain but is made ineligible for recommendation.
On X, the picture the report paints is one of near-abandonment. Since Elon Musk’s acquisition of the platform, the HCC notes, X has adopted a permissive approach to moderation alongside deep cuts to its trust and safety teams — leaving researchers and civil society with little confidence that reporting anti-Muslim content will lead to any meaningful action.
The report found that discussion of the mosque attack generated more than 112,000 mentions on X from 70,000 unique accounts, reaching an estimated 243 million accounts within six days.
Two-year pattern, not a single event
While the Talbot Street video is the report’s centrepiece, the HCC situates it within a longer study of online mobilisation.
Researchers analysed conversations on X over two years, from July 2024 to June 2026, using a commercial social-listening platform, and found that anti-Muslim discourse is not a spontaneous reaction to individual incidents but a sustained enterprise.
During periods of relative calm, the report found, influential accounts maintain a continuous background conversation centred on demographic change, “replacement,” cultural decline, housing, welfare, and public safety — building a shared framework of grievance before any crisis occurs.
When a local incident then happens, activists and influencers rapidly repackage it as proof of a wider threat from Muslims, Islam, or immigration, using emotive language, misinformation, and the selective use of facts to turn isolated events into stories about communities in decline.
The study also underlines the transnational character of the phenomenon. Well-established anti-Muslim networks in the United States and United Kingdom, supported by high-reach influencers and alternative media ecosystems, play a significant part in amplifying Irish events — and that international engagement, by signalling popularity to recommendation algorithms, feeds back into greater visibility for local audiences, deepening polarisation in a self-reinforcing loop.
“Ireland is now a source of content for international far-right influencers,” Malone said, warning that such amplification “should concern anyone responsible for the safety of our online spaces.”
Rhetoric spilling into the real world
The report’s authors place these findings against a documented history of anti-Muslim violence in Ireland. Between 2018 and 2025, the HCC identified 38 arson incidents and eight threats of arson linked to far-right mobilisation online and offline, Malone said, with anti-Muslim rhetoric frequently present in those cases.
He has described the tone of the hatred as increasingly violent, and its trajectory as spilling out of the online world into physical attacks.
“Islamophobia is very much the DNA of all far-right movements,” Malone said.
Islamophobia in Ireland, the report argues, came to prominence even before the broader wave of anti-immigrant mobilisation of the past two years — imported over the previous decade through a small but influential network of anti-Muslim activists with ties to European and British far-right movements, who used social media to spread narratives, mobilise supporters, and shape public debate.
What HCC wants
The report’s recommendations focus on the machinery of amplification rather than individual users. Beyond pressing Meta for an incident-specific risk assessment, the HCC backs a recommendation from the Oireachtas media committee to turn recommender systems off by default, and calls for stronger enforcement of Ireland’s Prohibition of Incitement to Hatred Act 1989 in response to incidents involving anti-Muslim hate and incitement to violence.
The underlying argument is that as long as outrage and hatred remain the most reliable route to virality, the incentives driving this kind of content will not change on their own.
As Malone put it, the public — and the Muslim community in particular — needs clarity on how Facebook’s recommender systems turn attacks on places of worship into viral content that normalises the event and calls for more violence.
We must take charge of our narrative. Research indicates that 92% of their Muslim coverage is negative. Help us change that.
Have Your Say!
Have you been affected by the issues raised in this story? Please share your experiences by emailing haveyoursay@soundvision.com.
Please include a contact number if you are willing to speak to a Muslim Network TV’s representative. You can also get in touch in the following ways:
Tweet: @MuslimNetworkTV
