Hillary Clinton’s surrogates and campaign aides faced what they called an overwhelming wave of Twitter attacks during the 2016 election cycle — some of it, they’re convinced, driven by Russian bots and trolls.
Now Senate Intelligence Committee investigators plan to ask Twitter representatives on Thursday whether false information spread by Russians accounts made it into real news stories aimed at torpedoing the Democrat’s candidacy, and whether the social media company took any steps to stem the fake news.
Story Continued Below
Thursday’s closed-door briefing in the Senate’s Hart office building is the latest move by federal investigators to probe social media’s role in last year’s presidential campaign, putting Twitter into the same hot seat occupied for weeks by its larger rival Facebook.
Facebook was a natural first target for congressional scrutiny, given its enormous 2 billion user reach and its role as an intermediary between trusted friends and family. But Twitter is also of keen interest to the Intelligence Committee because of the company’s outsize influence in distributing and amplifying news, a source close to the investigation told POLITICO.
Among other topics, the Senate staffers are eager to learn more about how disinformation posted on Twitter found its way into the news. They’re also interested in exploring whether automated Twitter accounts known as “bots” were used to boost the Google rankings of fake or misleading news items.
Senate investigators are inviting Google, Twitter and Facebook to testify publicly on Nov. 1 about attempts by Russia to use digital platforms to influence the presidential race, a Senate aide said Wednesday.
A former senior U.S. intelligence official said the Russian use of Twitter is “consistent” with what U.S. intelligence agencies found in their election post-mortem, which concluded that hackers associated with Vladimir Putin’s regime had interfered in the presidential campaign. That includes “the methodology of using multiple different lanes in an information operations campaign, » said the official, who spoke on the condition of anonymity to discuss details of the investigation.
Twitter’s power to influence the news cycle makes it a convenient tool, people who have studied the social network’s role in the election say.
“The fastest way to move a story from outlying media to mainstream media is to promote it on Twitter,” said Clint Watts, a counterterrorism expert and former FBI agent who has studied Russian influence campaigns. “All mainstream media is watching Twitter and it spreads organically from there to other platforms.”
Former Federal Election Commission Chairwoman Ann Ravel said tweets from foreign entities intended to cause trouble in U.S. elections should be no less alarming to lawmakers and regulators than the paid advertisements at the center of Facebook’s role in the Russia investigation.
« The sock puppets, the fake accounts cost money,” Ravel told POLITICO. “It’s not just a couple of guys in Moldova. If it were just a bunch of kids doing it for no reason, or a bunch of crazies putting stuff up on the internet, OK, that’s fine. That’s not political ads. But these are political ads: People are being paid to do it. »
Clinton supporters say they sensed early on that Twitter would be surprisingly toxic terrain.
“You’d get dogpiled for days if you expressed support for Clinton,” said one supporter, Brooklyn-based author Sady Doyle, who shared screenshots of critics who lambasted the Democratic hopeful as an “unindicted criminal,” “liar” and “fraud.” Doyle added: “One time I got something that was either a bot or a very bad impersonation of an American Trump supporter because it was a tweet in Cyrillic that when I translated was something like, ‘You are an ugly and bitter woman.' »
But it’s unclear how much help Twitter will offer the investigators.
Facebook showed up for its own Senate briefing earlier this month with a detailed story to tell about how Russian actors had used its site. The company shared seven out of about 3,000 ads linked to a St. Petersburg-based “troll farm” called the Internet Research Agency — a level of preparation the source close to the investigation called a “happy surprise.”
That raises the stakes for Twitter, which declined to say what it plans to show Senate staffers. A Twitter spokesperson has simply said the company is cooperating with congressional investigations.
But Senate investigators are coming to the debriefing with specific questions, the source said. Chief among them: What, if anything, did Twitter do to address bots once it had identified them as potentially malicious?
The prevalence of bots is a known problem on the platform, and researchers at the University of Southern California’s Information Sciences Institute found, in a study published on the eve of Election Day, that an estimated 400,000 bots were participating in the conversation around the election during a six-week period in September and October 2016. That represented about 19 percent of all election-related tweets.
The congressional investigators are also eager to learn more about how disinformation posted on Twitter bled into news outlets. This spring, researchers from Harvard and the Massachusetts Institute of Technology documented “a network of mutually-reinforcing hyper-partisan sites” including Breitbart and Infowars that gleaned content from Twitter.
Moreover, the Senate aides are interested in exploring whether the bot-driven amplification of tweets was used to boost the Google rankings of fake news. Google does not fully disclose how its algorithm works, but in 2010 a company official in charge of search-result quality revealed in a video posted on YouTube that the “organic rankings treat links the same from Twitter or Facebook” as they do from traditional websites. (Google periodically modifies its ranking methods without revealing the extent of the changes, so it’s unclear whether that practice has changed since then.)
U.S. intelligence officials have been closely studying Russia’s interest in tapping social media to fulfill its own geopolitical ends. A report on Russian election interference that the director of national intelligence issued Jan. 6 concluded that Russian elements enlisted social media trolls « as part of its influence efforts to denigrate Secretary Clinton. » Those efforts included pro-Kremlin online activists who prepared to mark Clinton’s expected victory by circulating the hashtag « #Democracy RIP. »
Even without foreign meddling, the internet might have remained a hostile place for Clinton. As a candidate, the former secretary of State has often been criticized for what’s judged to be a lack of authenticity, a quality often held in high esteem online. Former President Barack Obama openly joked about Clinton’s lack of online prowess, remarking that “Hillary trying to appeal to young voters is a little bit like your relative just signed up for Facebook.”
But Clinton herself has concluded that the overwhelmingly negative reception she got on the internet was ginned up at least in part by Russians. She dedicated four pages of her campaign memoir to the topic, writing that both Russian publications and independent trolls circulated stories saying she was « a murderer, money launderer, and secretly had Parkinson’s disease » — themes that often quickly found their way to right-wing outlets.
Twitter has long avoided drawing too much attention to the prevalence of bots on its platform. In part, that’s because such artificial accounts raise questions about the company’s already shaky business model. The platform makes money selling advertising in the form of promoted tweets, and shrinking its already stagnant user base by culling fake accounts would make it potentially less attractive to advertisers.
Twitter has defended its hands-off approach by saying it encourages the free exchange of ideas and information, and that journalists, media watchdogs and engaged citizens will push back on fake or manipulated tweets.
But the former senior intelligence official and other observers of the 2016 campaign say Twitter’s response doesn’t address the fact that a sophisticated bad actor — such as Russia — can create entirely new and false narratives by manipulating the size and scale of Twitter activity.
The cybersecurity firm FireEye reported that many hundreds of fake accounts with links to Russian entities were actively posting anti-Clinton messages, especially as Election Day neared.
“And then the day after the election, all of it decreased,” said Frank Cilluffo, a former Bush administration cybersecurity official who testifies frequently before Congress on the issue.
Much of the Russian effort to manipulate Twitter and Facebook was aimed at exploiting controversies over such issues as immigration and police shootings of minorities, according to Cilluffo, who directs the Center for Cyber and Homeland Security at George Washington University. “It was more that they were driving wedges into existing cracks in our society, around divisive issues, and trying to provide legitimacy to fringe ideas.”
“And that’s very consistent with the old KGB tradecraft, like [the false rumor] that the CIA created AIDS, but leveraging social media to do it,” he added. “And if it fails, it’s cheap.”
Elana Schor contributed to this report.