News

No, Russia and China Didn’t ‘Weaponize’ QAnon. It’s a Homegrown Nightmare.

A new report about QAnon makes a bombshell claim that data scientists say doesn't hold up.
The United States Capitol Building in Washington, D.C. was breached by thousands of protesters during a "Stop The Steal" rally in support of President Donald Trump on Jan. 6 2021 (zz/STRF/STAR MAX/IPx​)
The United States Capitol Building in Washington, D.C. was breached by thousands of protesters during a "Stop The Steal" rally in support of President Donald Trump on Jan. (zz/STRF/STAR MAX/IPx)
Logo_Disinfo Dispatch Padding
Unraveling viral disinformation and explaining where it came from, the harm it's causing, and what we should do about it.

A new report published this week makes the bombshell claim that the Russian and Chinese governments have helped fuel much of the QAnon activity we’ve seen over the last year. 

It’s an argument that sounds plausible. We already know that QAnon is a truly global phenomenon. And we know the Kremlin—and increasingly Beijing—put a lot of resources into spreading disinformation aimed at destabilizing western society, and the U.S. in particular. 

Advertisement

Many major news outlets ran with the story, using inflammatory headlines claiming Russia and China were in part responsible for QAnon and had “weaponized” the conspiracy theory for their own ends. 

But data scientists have slammed the report, claiming the conclusions it draws about the location of people posting about QAnon cannot be verified. Now, they’re warning that these unproven claims that QAnon is a foreign influence campaign undermine efforts to combat the real root cause of this homegrown extremist movement.

“It's what happens when non-data scientists have access to enterprise level tools and think that the end result is actually evidence,” one data scientist who didn’t want to be named because they fear professional consequences, told VICE News. “It's just bullshit because you can't demonstrate or prove that. Their data science is terrible and it's not a data-driven paper.”

The 45-page report was published by the Soufan Center, a well-regarded New York-based research firm focused on national security threats. The group’s researchers found that nearly one-fifth of 166,820 QAnon-related Facebook posts between January 2020 and the end of February 2021 originated from overseas administrators, the majority of them from Russia and China. 

But almost immediately, data scientists and researchers raised serious questions about the validity of the data shown in the report.

Advertisement

One of the central problems is that when anyone sets up a Facebook group, the person  self-assigns their location, and without access to the IP addresses of the accounts in question—and only Facebook has access to that information—there is no way that the Soufan Center could say for certain where these accounts are located.

As well as looking at “known foreign accounts” the researchers also assigned content “a foreign influence score.” This was based on feeding the content into a proprietary system created by a data analytics firm called Limbik.

This system has been trained to identify English-language content coming from certain countries, such as Russia, China, Iran and Saudia Arabia. To parse this content, Limbik uses a technique it describes as “linguistic anomaly detection.”

But data scientists who critiqued the report had never even heard of this term and didn’t know what it meant.

“What even is ‘linguistic anomaly detection’?” Fabio Votta, a Dutch data scientist asked on Twitter. “It's not a thing in natural language processing. Google gives 8 results, none of which seem relevant.”

Data scientists said that by using this method to identify foreign accounts, the Soufan Center was always going to return misleading results.

“Using techniques like this, a person in the U.S. who speaks English as a second language, for example, could be determined a foreign actor if the computer also deems their writing style more similar than not to content it knows has come from abroad,” Jordan Wildon, a disinformation researcher at Logically, an organization that identifies, monitors and combats the spread of misinformation, told VICE News.

Advertisement

“On the flip side, foreign actors who don’t post in a similar way to what the paper calls ‘known foreign content’ would not be identified.”

Another issue, raised by Marc Andre Argentino, a PhD candidate at Concordia University who studies QAnon and similar movements, is that the analysis is based on just 166,000 Facebook posts collected over a 13-month period to February 2021. 

Argentino said this represents just a tiny fraction of the total QAnon posts that appeared during that period, pointing out that a asmple he collected during January 2020 and August 2020 was ten times larger.  

“There were 1.65M posts in the 416 QAnon pages and Groups I was curating,” Argentino tweeted. “Considering they claim a full year of collection I do not think their sample is large enough.”

QAnon is a conspiracy theory that originated on the message board 4chan and is centered around U.S. politics. The vast majority of adherents to QAnon are based in the U.S., but from the very beginning the conspiracy movement has had an international presence. 

Now, fueled by the pandemic, QAnon’s global footprint has grown dramatically in the last year. One researcher has identified QAnon groups in at least 85 countries.

So it’s no surprise that QAnon content is being posted from accounts based outside of the U.S. But that does not mean that this content is state-sponsored disinformation.

“Even if a post comes from another country, that doesn’t mean it’s state-sponsored influence, as sensationalist headlines around the report have implied,” Wildon said. “Using the same methodology as that paper, research I’ve conducted into QAnon could identify more than 30 countries exerting ‘foreign influence’ when in reality the posts are simply from QAnon adherents living in different countries.”

Advertisement

Part of the problem, Wildon says, is that “the paper doesn’t explain its methodology, or what it defines as ‘known foreign content’ in any detail, and it would be almost impossible to replicate, and therefore verify, these findings independently.”

The Soufan Center says it stands by its research. “We have full confidence in the data and analysis presented in our report,” a spokesperson told VICE News, adding that the system relies on more than just location data and analysing the text of social media posts.

But this stance has already been undermined by the founder of Limbik, Zach Schwitzky, who acknowledged in an interview with Yahoo News that identifying foreign content was not “an exact science” since there is rarely publicly available account information about individuals or groups who post on Facebook.

Facebook did not respond to VICE News’ questions about whether the Soufan Center’s analysis was accurate.

The report’s findings have been twisted by certain groups to help bolster the unproven claim that Russia helped create QAnon. This conspiracy theory about a conspiracy theory has been circulating for some time, but has failed to gain significant traction. 

But the report could have an even greater impact on the fight against QAnon.

The Soufan Center, which was founded by former FBI agent Ali Soufan, is well connected inside the U.S. government, meaning that decision makers could take notice of the report and potentially redirect funds from investigating QAnon’s roots inside the country and put them into fighting QAnon as part of the fight against foreign disinformation campaigns.

“Decision makers are gonna watch this and they're going to say, 'Oh shit it's China and Russia,'” the data scientist who works in this area told VICE News.

“So you're going to reallocate resources from your domestic threat assessment actors to determine whether it's violent extremism or not to, ‘Oh, it's foreign interference.’ So you're taking the problem outside of your country, blaming others, and ignoring the core root causes. You're basically sweeping under the rug.”