Elon Musk’s hellsite isn’t making us fall for Pierre Poilievre. It’s doing something far scarier

The Dead Internet Theory, as it became known, has been dismissed as conspiracy and lauded as biting satire — a commentary on an information superhighway so overrun with bad actors and cranks that it feels like all the real humans fled long ago. But maybe the theory wasn’t so absurd after all. Two recent Canadian political stories suggest big parts of the internet have, in fact, become zombified graveyards. What’s worse, Canada’s political and media elite are now allowing the most toxic plot in the whole cemetery to infect the real world.

Twitter, the long-standing clubhouse of Canada’s media and political elite, has become a haven for conspiracy, disinformation and rage, particularly since it was purchased by Elon Musk in 2022 and converted into X. The site is now replete with trolls, culture warriors and bots. And yet most of Canada’s top journalists and politicians continue to stick around, letting all that toxicity bleed into our national conversation. 

These stories should be a jarring wake up call. They show how an internet of decaying platforms, especially but not only Twitter, has made us irrationally adversarial, quick to delegitimize and demonize our political opponents. They should be an invitation, at the very least, to unplug from Musk’s nuclear tailing pond. While I once loved the cut-and-thrust of Twitter, I noticed its decline into madness last year and decided to log off. For the sake of our collective sanity, it’s time the rest of Canada’s political and media class did the same.

The story begins earlier this month, when a Twitter user named @The280Times noticed some fellow tweeters from suspicious locations buzzing about the exact same thing.

There was Mica, from Jodhpur, India; Laurette, claiming to be from Hawaii; and Conception, who listed their hometown as both Manchester and Salt Rock, South Africa. All posted the same message, all within minutes of each other: “Just returned from Pierre Poilievre’s rally in Kirkland Lake and I’m still buzzing from the energy!”

All told, hundreds of accounts posted iterations of that message, one after another, on Aug. 3, each writing: “It’s refreshing to see a leader who actually listens to our concerns.” @The280Times, not unreasonably, declared this a case of “potential foreign interference.” These were no Poilievre fans, they suspected, but automated accounts programmed to parrot the same bit of propaganda.

Just a day earlier, a different story with virtually identical origins popped up on Iran International, a Saudi-linked news agency critical of the Iranian regime. Citing a new report from an Israeli firm, XPOZ, the story claimed that Iran had “masterminded anti-Israel protest in (a) Canadian university.”

The Iranians did it, XPOZ said, with bots. It was “a massive activity, funded, co-ordinated and organized by a foreign government,” the organization told the news network, to trick people into joining the pro-Palestinian encampment at McGill University.

After these stories emerged, recriminations flew. 

NDP Member of Parliament Charlie Angus took to Twitter, the scene of the crime, to ask: “Why are Russian and American bot accounts interfering with Canadian politics on the eve of an election?” The MP raised the possibility that it was the Poilievre campaign itself using these “offshore botfarms.” The Conservatives were pressed by numerous outlets, including this one, to deny they were responsible.

Meanwhile, Neil Oberman, a Montreal lawyer who represented a coalition of Jewish students trying to obtain an injunction against the McGill protest, weighed in on the Iran “scoop.” He told the National Post the XPOZ study should be a clarion call for Canadians. “Canada cannot be allowed to be manipulated by foreign countries to create social upheaval,” Oberman, who is now running for the Conservatives, said. “That is not acceptable. Period.”

But I’m here to tell you that it’s not bots manipulating or creating social upheaval: It’s Twitter itself. These bots are no more capable of convincing students to care about the abhorrent loss of civilian life in Palestine than they are of tricking voters into getting excited about Poilievre. And the more time we spend ascribing the behaviour of our fellow citizens to automatons on social media, the more polarized and paranoid we will become.

The truth is many Canadians, a clear plurality of them at this point, do like Poilievre. Most are, indeed, buzzing — just not Mica from Jodhpur. At the same time, the students who camped out for Gaza at McGill, and other campuses across Canada, weren’t there because they were conned by Iranian bots posting in Farsi, but out of genuine belief for the plight of the Palestinian people. The only people who seem sure of Twitter’s brainwashing potential are those who spend far too much time on Elon Musk’s hellsite.

“Twitter is a dumpster fire,” Fenwick McKelvey, an associate professor at Concordia University, told me last week. “Anyone talking about Twitter being a credible political space, I take skeptically.”

After the 2016 U.S. election, when bot paranoia was in full swing, McKelvey undertook a sprawling investigation into how inauthentic social media users could interfere with Canadian democracy — or, perhaps, how they were already at it. But his research, published in 2019, didn’t find much fire beyond the smoke. The threat of bots was “overstated,” he wrote. (Full disclosure: McKelvey interviewed me for his project.)

What McKelvey discovered instead is something more damning about our political culture. “Bots became a way of policing who is an authentic and inauthentic political actor,” he told me. Accusing your political opponents of employing bots — or being bots themselves — was an easy way to discredit everything they believe in. This was much easier if those users spoke languages other than English and French. That pattern gave McKelvey “hints of bad vibes and xenophobia.”

Botnets can be created or acquired with relatively little technical knowledge and about $100. As a result, they’ve become increasingly common. Sometimes it can be difficult to tell if a network of bots has a particular aim or goal, or if they are merely floor models: On display for a bot-master to sell to a prospective client. The pro-Poilievre bots, for example, were also tweeting about the Olympics and in support of climate action. 

The more I looked at the coverage of the XPOZ report, the more McKelvey’s warnings rang true. I asked the company, repeatedly, for a copy of the report itself — but a representative refused to hand it over. Nor would XPOZ provide any detailed information about the company’s methodology, confirm how many tweets were supposedly sent by this malign Iranian propaganda campaign, or tell me how impactful the supposed botnet actually was. This is highly unusual for a firm ostensibly trying to combat disinformation.

All XPOZ would say is that they scraped publicly available tweets, made their best guess about where the accounts were based, and then used their “proprietary model to establish inauthenticity.” I can’t confirm or validate their work because they refuse to show it. 

“It’s a pretty incredulous way of dismissing an entire student movement,” McKelvey observed. At most, the company may have proven that Iran helped amplify some of these pro-encampment messages. “You can amplify all the messages you want, it’s not going to get me to camp out in the rain,” he laughed.

As for those pro-Poilievre bots? “What I find funny is how sloppy this is,” McKelvey said. There is no evidence linking them to the Conservative Party, nothing even to suggest the party would have gained anything real by wasting time or money on an amateur-hour botnet campaign. 

And yet, both of these stories became stories anyway. They went viral on Twitter then spurred mainstream news coverage — two evidence-free red herrings that did nothing but convince even more Canadians that anyone who disagrees with them is devoid of independent thought and free will.

The strangest thing about all of this is that bot campaigns don’t seem to work.

When Russia dispatched an army of bots and trolls to muck about in the 2016 presidential race, the novelty of the attack set off alarm bells. But a retrospective 2021 study, published in Nature Communications, found “no evidence of a meaningful relationship between exposure to the Russian foreign influence campaign and changes in attitudes, polarization, or voting behaviour.” (That doesn’t, of course, let Moscow off the hook for trying.) 

The constant accusations of “bot!” have probably done far more damage than the bots themselves. “It’s not as though Poilievre supporters are going to be like ‘oh he’s not as popular as I thought he was,’” McKelvey says. “It’s going to be: ‘The liberal media diminishing my lived experience of liking this guy.’”

The real problem of Twitter today isn’t bots, or even the weirdo billionaire running it, it’s the fact that otherwise credible journalists, pundits and politicos keep treating the platform as a legitimate and useful forum for democratic discourse. It’s not.

Even in our era of acute political and social polarization, Twitter stands out. It is an exceptional hall of mirrors, distorting and reflecting back our political divides in wonky and dangerous new ways. It’s not the only one, of course: Facebook, we know from insider leaks, let hate speech flourish and encouraged rage-inducing content to make money. 

But Facebook at least pretended to address the problem. Musk is egging it on.

His stated goals, when he bought Twitter, to encourage free speech and vigorous dialogue, were perfectly laudable. The problem is the site, both through its mechanics and culture, isn’t set up to be a productive forum. It actively rewards and promotes squabbling, invented nonsense, anti-social behaviour and even the bots themselves. 

The fact that so many politicians and reporters continue to spend their days on Twitter means the general public is now catching second-hand brain worms. That kind of seepage leaves ordinary Canadians, even if they aren’t personally hooked on Twitter, believing Iranian androids are living in tents at McGill, or convinced Poilievre is delivering his stump speeches to rooms full of holograms.

We, as a society, and especially those of us in politics and journalism, made a conscious decision to let Twitter — and Facebook, Instagram, Google and more recently, TikTok — inform our political lives. It was a choice made long before the internet died. (I know, I was there.) At first it was exciting. We were revolutionizing how news is reported and how politics get discussed, democratizing the flow of information.

We believed the tech executives behind these sites were more moral and knowledgeable than the rest of us, committed to those same goals. But it’s plain now they were never motivated by anything other than greed, glory, and advertising revenue. Today, the digital infrastructure we use to manage our national conversation is warping, distorting and bastardizing our public square in stranger and scarier ways than we’ve yet come to appreciate. 

Some, including Prime Minister Justin Trudeau, believe we can fix this problem with more energetic regulation. It’s a dicey proposition, one entirely undercut by the fact that Trudeau himself still frequently posts to Twitter, which has thumbed its nose at regulators everywhere. That’s to say nothing about Poilievre, who may as well be the Member of Parliament for Twitter.

If we want to free ourselves from the bots and the shadowboxing that comes with them, we should start by logging off that hellsite, at the very least. More than anywhere else, it has become a fount for misinformation, conspiracy theories and bad-faith political chicanery — all cheered on by its billionaire owner. 

“The more attention we give it,” McKelvey says, “the worse our politics are.”