After laboring in secret for months, a new report commissioned by the Senate Intelligence Committee names a culprit more useful on a user-for-user basis to Russian espionage services than Facebook or Twitter: Instagram.
The Senate committee ordered the report from researchers at cybersecurity company New Knowledge; Jonathan Albright, director of the Tow Center for Digital Journalism Digital Forensics Initiative (part of Columbia Journalism School); and Canfield Research’s Ben Johnson. The group examined data supplied to the committee by Facebook, Twitter, and Google. Albright says that he hopes this puts an end to the war between the company and the media organizations reporting on it—the US government has weighed in, using data supplied by the companies themselves, and has vindicated claims made by dogged reporters that Facebook often contradicted.
It’s one of two reports to the Intelligence Committee released on Monday, the other coming from Oxford University and private network analysis firm Graphika. The Oxford/Graphika report emphasizes the nature of Russian social media messaging to prospective voters, focusing on infighting on the left and unifying issues on the right.
The New Knowledge report carves into stone much of what has been reported and rumored about the Russian military’s attack on the 2016 election, using data from the tech companies themselves. Albright says that Instagram’s relative absence from investigations into Russian interference is not a measure of its importance to the Russian GRU security service during 2016, but of the company’s success at triaging brand damage to its younger, more promising product. With this report, he hopes that will change.
“They’ve taken great care to keep press mentions of Instagram out,” Albright observed. “Even on the Facebook datasets that I put out, they didn’t admit to [Russian use of] Instagram until days later. It’s very nice to see a more official data set that they can’t walk back from.”
But how successful exactly was Instagram by comparison to its much larger sibling and its notoriously troll-ridden competitor, Twitter? “It saw more success than either Twitter or Facebook combined,” Albright says bluntly. “It has all the infrastructure of Facebook’s advertising behind it. It has Facebook’s sophisticated targeted audiences behind it. And it’s hard to research and messy. It actually functions as information distribution more like Twitter—it has most of the power of Twitter and a lot less of the transparency.”
The Russian Instagram accounts run by the troll farm known as the Internet Research Agency (IRA) appeared before the IRA’s Instagram accounts, Albright says. “They lasted longer than the Facebook accounts, they saw much, much more engagement—in the hundreds of millions more.” For example, one account, @feminism_tag, ran left-ish anti-Hillary memes from a supposedly “intersectional feminist” perspective according to the report; Albright said it had no equivalents, in terms of penetration and engagement, on Facebook or Twitter.
“Facebook had 61,000 posts and Instagram had over 100,000 posts [in the data sets provided to the researchers], and when you look at the volume on Twitter, it’s like 10 million,” Albright said. “But Instagram still won. There were thousands of Twitter accounts that had no engagement; they were just spamming. But the Instagram stuff—you couldn’t have called it spam. They saw double the number of likes as Facebook, and more comments.”
It’s taken two years for the postmortem on election interference in 2016 to reach a point of public certainty about the Russian government’s efforts to elect Donald Trump. This is in no small part because Facebook deleted vital information about Russian tampering with its service when Albright began researching it in October 2017. The company also pushed back regularly in public statements later proven false. Even without active resistance by a company worried about its bottom line, “the ability to do good, accurate network analysis is going to be really hard,” Albright says.
The new report, Albright says, is distinguished by something that pleases him: It’s very hard to contradict. “It’s based on data,” he tells CJR. “It’s very nice to see a more official data set that [Instagram] can’t walk back from.”