October 25, 2026
Over the last decade, politicians, NGOs, and journalists across the West have blamed online speech for real-world political violence. In Australia recently, politicians and others blamed social media for an attack on a left-wing member of parliament, Helen Haines, who was run over by a man in his car. “We’ve got these big US social media companies and their business model, to me, seems pretty incompatible,” said Senator David Pocock after the attack. “That’s something we’ve got to look very hard at as a country and as elected representatives.” Said Bill Shorten, a former Labor leader who now runs the University of Canberra, “I think that some people, through the algorithms of the big tech bros and social media, have been radicalized online.” Mike Burgess, director-general of the Australian Security Intelligence Organisation (ASIO), Australia’s FBI, decried a “permissive environment of violence.” And Commissioner of the Australian Federal Police Krissy Barrett said the things politicians said online “does have a direct correlation to what we see in terms of threats and abuse.”
But it was clear within hours that the alleged attacker of the Australian MP was intoxicated and likely mentally ill, and within days it was revealed that he and his family had been friends with her for years. A source close to the investigation said he “was known to police” and that “It is understood he has undergone a mental health assessment.” People online quickly discovered that the alleged attacker had posted regularly about climate change and Gaza and that the head of the Australian Green Party followed him on X. As such, he attacked the member of Parliament closest to his apparent views, which challenges the notion that his politics motivated him. As for Barrett, she declined to give a single example of things people said online resulting in real-world violence.
Correlation is not causation. Killers may have consumed violent media, but so too did practically everyone else in society. It would make sense that would-be killers, who often fantasize about their violence before doing it, would consume more violent content. And so if there is a relationship, it could show the opposite causal direction.
Moreover, relying on the testimony of killers, many of whom suffer from personality disorders or mental illness, is irrational. The man who shot Ronald Reagan said he did so out of an obsession with Jodie Foster in Martin Scorsese’s “Taxi Driver.” Nobody proposed censoring “Taxi Driver,” Scorsese movies, or Foster, and yet politicians and bureaucrats around the world proposed mass online censorship because the Buffalo shooter said he was inspired by Christchurch.
The response to this from those advocating greater mass censorship of social and other forms of media is that the killers themselves often say they were influenced by media, and scholars can document copycat killings that people hear about through media.
For example, a Buffalo shooter in 2022 cited a Christchurch video of a mass shooting. New York Attorney General published a report on a Buffalo, New York mass shooting, which found that the killer’s radicalization “began upon viewing on the 4chan website a brief clip of a mass shooting at a mosque in Christchurch” and concluded, definitively, “Anonymous, virtually unmoderated websites and platforms radicalized the shooter.”
Twelve days after the Buffalo shooting, Nina Jankowicz, who in 2022 briefly led the Department of Homeland Security’s “Disinformation Governance Board,” said on National Public Radio’s (NPR) Fresh Air in 2022, “Violent extremism is begotten by things people see on the internet.”
Censorship advocates appear to find support from scholars. “There is always a copycat risk with this kind of graphic material,” said Julie Inman-Grant, Australia’s top censor. Karsten Müller and Carlo Schwarz, two economists writing in the Journal of the European Economic Association in 2021, studied 3,335 anti-refugee incidents and found that “a one standard deviation higher right-wing social media usage is associated with a 10% higher probability” of an attack in a municipality. During internet outages, they found, the probability of an attack on refugees fell 53% relative to the mean. A Russian study by Leonardo Bursztyn and three colleagues found that a 10% rise in penetration of the social network VK “increased hate crimes by 21.7% in cities where [nationalist political party] Rodina received the most votes.”
But the New York report conceded that “the full progression of his radicalization is not known to be documented,” and authors of the studies drew much narrower conclusions. “We do not claim that social media itself causes crimes against refugees out of thin air,” wrote Müller and Schwarz. They described social media as “a propagating mechanism for hateful sentiments” where hostility already existed. The Russian effect appeared only in nationalist cities. In cities where Rodina polled poorly, Bursztyn and his coauthors found “no effect,” and “we do not see a significant effect of social media penetration on self-reported hostility.”
The largest experiment testing this link used Facebook data during the 2020 election. Brendan Nyhan and colleagues, writing in Nature in 2023, found that reducing exposure to like-minded sources “had no measurable effects on eight preregistered attitudinal measures…such as affective polarization, ideological extremity, candidate evaluations and belief in false claims.” A January 2025 Campbell Systematic Review of 55 studies on exposure to hate in media, rested its estimate for offline violent behavior on just two studies.
And there are many good reasons to distrust what killers claim their motives were. We do not trust killers to control their own behavior, which is part of the reasons why we incarcerate them. We do not trust killers to tell the truth in general. And yet the would-be censors would like us to trust their account of why they did what they did, even when they are plainly mentally ill or, as in the case of the Australian man, also intoxicated.
In truth, all of us, killers and non-killers, are famously “strangers to ourselves,” philosophers have found for millennia and psychologists have found for decades. We often don’t know why we do what we do and then make up reasons after the fact, as many studies and real-world observations find.
There is even some evidence that media can decrease real-world violence simply by taking up time that might be spent committing crimes. A 2009 study that matched theater audiences to police reports from 1995 to 2004 found that “violent crime decreases on days with larger theater audiences for violent movies,” so that “violent movies deter almost 1,000 assaults on an average weekend.” In 2016, found researchers, “We find no evidence of an increase in crime associated with video games and perhaps a decrease.”
In truth, people have been blaming media for violence for 140 years, and courts in the US and across the West have for decades consistently rejected any link, which is why they have disallowed much government censorship. In 2011, the US Supreme Court struck down, 7-2, California’s video game law. “These studies have been rejected by every court to consider them,” wrote Justice Antonin Scalia for the majority, “and with good reason: They do not prove that violent video games cause minors to act aggressively.”
And yet educated and intelligent people across the Western world continue to make the link. Why is that? And what can be done to counter it?
Political and psychological motivations
Many appear to use real-world violence as an excuse for greater government censorship that they want for other reasons. Such was transparently the case in Australia recently, where the politicians claiming a link between the attack on a politician and social media had already come out for government censorship proposals.
Moreover, the liberal globalist establishment since 2016 has successfully sought to censor right-wing populist movements in particular. We and others have documented the roles played by liberal and “never-Trump” censorship advocates within the Intelligence Community (IC), IC intermediaries such as NGOs and university research centers, in politics, in journalism, and on the wider Left. We have also documented how censorship advocates use many different justifications for censorship, including stopping misinformation, stopping hatred, preventing election interference, protecting women and minorities, and protecting children.
Much of this was clearly a reaction to the populist revolts in the UK in the form of Brexit and the US in the form of Trump’s election, albeit done in the name of preventing supposed Russian information warfare. Germany’s Network Enforcement Act, the NetzDG, which obliges platforms to remove illegal posts under threat of fines, said explicitly that it was reacting to “the experience of the US election campaign.” Germany’s parliament, the Bundestag, passed the law in 2017. Three months later the Alternative für Deutschland (AfD) rose in that election from 4.7% to 13% and entered the Bundestag for the first time with 94 seats.
The demands from people close to the IC for censorship intensified after 2016. The Aspen Institute’s Commission on Information Disorder, co-chaired by the television journalist Katie Couric, the cybersecurity expert Chris Krebs and the civil rights leader Rashad Robinson, asserted real-world harms from disinformation throughout its November 2021 report
But it gave as sources for its violence and hate claims various articles in liberal publications like The Verge, The Guardian and The New York Times rather than to scientific studies. “Yet we don’t know as much as we should about how, when, or where,” the report conceded.
In the US, censorship leaders sought control of social media platforms in the name of preventing radicalization, supposedly a precursor to real-world violence. “Young men were radicalizing because of content they were seeing on YouTube and Twitter,” Renee DiResta said in a pro-censorship PBS “Frontline” segment in 2018. And yet, when pressed, DiResta all but admitted she had no evidence. “Is it the platforms that are causing it, or are they just reflecting it?” she said, adding that “this is something that we still don’t have a very good handle on as outside researchers.”
That didn’t stop DiResta from forming the censorship-by-proxy effort overseen by the Department of Homeland Security and involving her Stanford Internet Observatory (SIO), which flagged posts and successfully demanded that social media platforms censor them, first on the 2020 election and then on Covid in 2021. DiResta said in a video for DHS that the project was designed to get around the First Amendment.
Of course, people claimed a link between media and behavior and demanded censorship long before 2016, and one theme running through the cases is that those making the link promote a view of human persuasion that runs opposite to the scientific consensus. The research, summarized by Princeton professor Hugo Mercer in his 2020 book, Not Born Yesterday, shows that people are far more stubborn in holding on to their views, including inaccurate ones, than they are malleable.
Mercer notes that the narrative told by censorship advocates, that people can easily mislead a gullible public, is not supported by research. And yet panic-mongers prey on the opposite perception. They assume people are blank slates, but in reality we carry natural and learned biases and assumptions that are difficult to change.
Consider the panic over dime novels, cheap and often lurid paperbacks that became popular in the late 19th Century. “Four loaded revolvers were found in their possession, and a quantity of these foolish stories,” wrote Anthony Comstock, a New York anti-vice crusader, of three runaway boys arrested in New York, in his 1883 book, “Traps for the Young.” He suggested that mere possession of the novels proved their causal power, just as DiResta asserted that young men were becoming radicalized simply because they were watching YouTube videos.
Many of the people claiming a link were progressive do-gooders. Next came films and comic books, and the accusations continued for decades. Jane Addams, the founder of Chicago’s Hull-House settlement, described in 1909 three Chicago boys who planned for weeks to lasso, murder and rob a milkman after seeing frontier adventures on screen.
It is worth considering that many of the people who see a link are themselves personally revolted and horrified by the violence they see depicted, or the online speech, and rationalizing a fundamentally irrational or nonrational and emotional desire to silence speech and image that makes them personally uncomfortable, often for understandable reasons.
In other words, while liberal elites certainly had a rational, self-interested motivation for the censorship they wielded, they also helped create a public backlash that helped their populist enemies, returned Trump to the White House, and delegitimized their censorship project. And so one has to consider that elites were driven by something far more base and crude, namely a pathological need to control what they should not control, namely the speech, thoughts, and behaviors of others.
As they spread hoax after hoax about Russian interference, “trans genocide,” climate change apocalypse, mass migration, Covid, and race (e.g., about police killings of whites vs blacks), progressives demanded mass censorship of accurate information and legitimate opinions about Russians, trans, climate, migration, Covid, and race.
Many of the leading advocates for censorship, from Hillary Clinton to Bill Gates, are arrogant and dogmatic in their certainty. They are intolerant of populist voices to the point of being hostile and aggressive, which only makes populists hate them more, and write more negative things about them online, which then further feeds their demand for censorship. Clinton was a censorship advocate long before the populist revolts. She introduced the Family Entertainment Protection Act in 2005 to bar sales of restricted-rating games to anyone under 17 and let the FTC run secret annual audits of stores.
Upset at being publicly criticized and humiliated, progressives including Australian Labor Party politicians have taken to scapegoating “the algorithm,” as though there is a technical reason for their unpopularity. Because they believe they are enlightened, and yet find themselves hated online, they seek not to persuade but to persecute.
The alternative for many of these secular liberals is that their pseudoreligious politics could be wrong, which would shatter their identities and dream worlds. Maybe men can’t become women, they might conclude. Perhaps mass migration isn’t great for working people and national cultures. Or possibly hysterical climate policies are bad for the economy. If those things were true, then what these politicians have been advocating for decades was at best intellectually wrong and worse, morally wrong.
It is notable that such politicians are seeking not just censorship but larger sums of taxpayer funding to finance government propaganda, and stronger demands on social media platforms to carry it. The EU, Canada, Australia, and nations across the West already pour millions into news organizations and are seeking to spend even more. German Chancellor Friedrich Merz is proposing to require social media platforms to make 20% of their algorithmic recommendations come from government-approved news media content.
Australia’s government is seeking the right to run fake accounts to “disseminate” information for “educational” and “promotional purposes,” just two years after demanding social media platforms crack down on fake accounts.
It is not surprising that the demand for greater control over social media comes from the same elites who have been attempting to control practically every aspect of society including the sex and gender of children, food, medicines, housing, transport, medical care, etc.
It is also part of a broader effort to punish populists. After Canadian truckers protested Covid restrictions, then Prime Minister Justin Trudeau intimated they were Nazis and froze many of their bank accounts. “Freedom of expression, assembly and association are cornerstones of democracy,” he said, “but Nazi symbolism, racist imagery, and desecration of war memorials are not.” In Australia, the Labor government has demonized millions of Australians as Nazis and racists for protesting mass migration after two Muslims murdered 15 people, most Jews, at Bondi Beach last year.
Having let go of traditional faith, many of these people behave as though they have exorcised all evil from within. In reality, they project their own dark thoughts onto millions of good people. Everyone who disagrees with them, they imagine, is evil and must be silenced in one way or another. It’s why progressives cut off friendships with ideological dissenters.
These individuals are so caught up in their self-righteousness that they do not believe they can be censored. In Australia, two Labor Party staffers appeared insulted when I warned that they could be censored by their political enemies one day and asked how that would make them feel. It was clear to them that censorship was only for bad people and that they were good.
This view reminds us that many censors have a weak “theory of mind” about the public and their opponents. They do not appear to understand that demanding censorship makes people distrust and even hate them. They suffer from the “third-person effect,” a psychological bias in which people believe disinformation affects others more than themselves, which is what Mercer criticized.
Progressives were similarly surprised when the public voted for Trump in part as a way to protect themselves from progressive persecution, whether through censorship, lawfare, or some other means.
Defeating Fake Linkages
Conservatives and Republicans are not immune to making facile connections between online and real-world violence. After an El Paso, Texas, mass shooting in 2019, then-House Minority Leader Kevin McCarthy cited unnamed research on Fox News to claim, “We’ve watched from studies shown before of what it does to individuals.” President Donald Trump blamed “gruesome and grisly video games that are now commonplace.”
All of this is troubling because killers have a strong self-interest in blaming their hideous behavior on some external factor and depicting themselves as victims of society, the media, their families, or other forces. When politicians, security officials, academics, and journalists support such claims, they risk reinforcing criminals' belief that they are not responsible for their actions.
The historical record also shows that when serious academics get involved in the research, they help debunk any link.
Even modest claims have been debunked. Britain’s National Council of Public Morals convened a Cinema Commission of Inquiry that concluded in 1917 that, while a general link between cinema and the rise in juvenile crime had not been proved. “The cinema suggests the form of the activity rather than provides the impulse to it.” But Cyril Burt, a London psychologist, debunked even that tenuous claim. He went through case files of London delinquents for his 1925 book, The Young Delinquent, and found that he was “convinced, from a careful sifting of each conceivable case, that direct imitation is exceptional and rare.”
Private philanthropists can play a positive role in commissioning studies. The Payne Fund financed 12 research projects between 1929 and 1933, published from 1933 to 1936, that debunked claims that motion pictures affected children and adolescents.
In 1933, Herbert Blumer and Philip Hauser, two University of Chicago sociologists, collected inmates’ own accounts with no comparison group of non-offenders and found films “a factor of importance in the delinquent or criminal careers of about 10 per cent of the male and 25 per cent of the female offenders studied.” Henry James Forman, a journalist hired to popularize their studies, claimed that 406 crimes committed in 115 films had served as a “university of crime” for children.
But, W. W. Charters, the Ohio State University professor who chaired the Payne Fund Studies, of which the Blumer and Houser study was one of 12, wrote that, “To say that the movies are solely responsible for anti-social conduct, delinquency, or crime is not valid.” Agreed Paul Cressey, himself a Payne Fund researcher, in the American Sociological Review in 1938, “Social causation is entirely too complex a problem to be explained by any such simplistic interpretation of incomplete data.”
Fredric Wertham, a New York psychiatrist who examined violent children, testified before a Senate subcommittee on April 21, 1954 that “comic books are an important contributing factor in many cases of juvenile delinquency….Hitler was a beginner compared to the comic-book industry.”
But Frederic Thrasher, a sociologist, wrote in 1949 that no acceptable evidence had been produced by Wertham or anyone else. And the executive assistant to the American Psychological Association (APA) told Senate staffers in a letter, the day before Wertham appeared, that, in his view, the research “does not permit inferences as to cause and effect relationships.”
In the end, the Senate subcommittee that heard from Wertham wrote that expert opinion held that it was “unlikely that the reading of crime and horror comics would lead to delinquency in a well-adjusted and normally law-abiding child.”
In 2012, Carol Tilley, a University of Illinois information scientist, went through Wertham’s case files and found that he had “misrepresented and altered children’s own words about comics.”
Close interrogation of experts is also essential. Wertham, for example, conceded in his own book that “never has a single child ever told me as an excuse for a delinquency or for misbehavior that comic books were to blame.”
The Australia case shows that pushing back aggressively as soon as censorship advocates assert a link appears to have an impact. I did so within hours, with my posts on X shared by over 1,000 individuals, including political leaders in Australia, who made similar arguments to the ones above.
Though the new censorship legislation could still pass, the same government’s misinformation bill failed in 2024 after opposition rose. The Institute for Public Affairs, other researchers, journalists, free speech advocates, and I have all successfully challenged the government and changed the conversation to focus squarely on censorship.
