Social media companies are struggling to contain new forms of political misinformation on their platforms that bubbled up during this year’s midterm elections.
The more unorganized, guerilla misinformation being spread is distinct from the foreign election interference efforts that dominated during the 2016 elections, and often starts from within the United States’ own borders, researchers say.
Jonathan Albright, a researcher at the Columbia University’s Tow Center for Digital Journalism, said that false and misleading information is still running rampant on platforms like Facebook.
“Many of the dangers that were pointed out years ago have seemed to grow exponentially on Facebook, not unlike the other large social media ‘platforms,’ ” he wrote earlier this month.
Albright analyzed 250,000 Facebook posts, 5,000 political ads and historic engagement metrics for hundreds of pages and groups on the social media platform.
“The takeaway: It’s not good,” he wrote.
Researchers point to a number of false or unsubstantiated claims that ran rampant on social media in the lead-up to the midterms this year, including posts baselessly accusing Democratic donor and hedge fund billionaire George Soros of funding the migrant caravan of Central American asylum-seekers heading toward the U.S. border and incorrect conspiracy theories about the origins of bombs sent to high-profile Democrats.
Misinformation also went viral on Election Day, including false videos of voting machines changing votes and false Facebook posts about Black Panthers “armed with assault weapons roaming the streets and neighborhoods as well as the voting polls in Georgia to intimidate” and “persuade” voters to support Democratic gubernatorial candidate Stacey Abrams.
The spread of such information appeared to mark a shift from 2016 in terms of the source of the content. While U.S. officials and researchers expressed concerns about misinformation campaigns backed by foreign governments in 2016, much of the hoax information that spread on social media during this election did not have clear links across multiple platforms.
“We’re at the stage where disinformation has been pushed out so many times that other people know how to do it well,” said Alan Rosenblatt, director of digital research at the left-leaning opinion firm Lake Research Partners.
Department of Homeland Security officials have said that they had not seen evidence of coordinated hacks or attacks on election infrastructure in the 2018 midterms, and social media researchers and the companies themselves have not yet uncovered any large-scale, nation-state coordinated misinformation campaigns tied to the midterms.
Rosenblatt compared the evolution of distributed misinformation to how changing tactics by extremist groups — from carrying out organized strikes to lone-wolf attacks — forced national security officials to change their approach to combatting terrorism.
Instead of working from an office in Russia to spread hoaxes, experts say, misinformation runs through different Facebook groups of people of similar ideologies. Misleading stories grow in strength as they bounce around within a group and then get passed to others before they eventually hit the mainstream.
Facebook is far from the only home for political misinformation online, researchers say, with Twitter and YouTube also working to combat hoaxes on their platforms. Albright has highlighted how Instagram has also become one of the worst platforms in terms of how it addresses conspiracy theories and hoaxes, which The Hill has reported.
As an example of how such information spreads, the conspiracy of Soros being behind the caravan of Central American migrants was pushed through fringe Facebook groups before catching steam on Twitter and eventually being raised by certain GOP lawmakers and President Trump.
Political observers note that many individuals spreading misinformation are often simply sharing content aimed at bending discourse to the most advantageous political ends, even if they are not tied to any type of broader, coordinated misinformation efforts.
People sharing misleading information also walk a fine line that becomes difficult for social media platforms to enforce, experts note.
Instead of outright falsehoods that ran rampant during the 2016 elections, individuals often share hyperpartisan and sensationalized stories that may not be entirely accurate but still contain some kernel of truth. And because they’re often not part of explicit foreign manipulation campaigns, they don’t clearly violate platforms’ terms of use.
“People who don’t have the access or ability to verify information, it just gets spread too fast for it to be contained,” Rosenblatt said. “I think it often spreads so fast that by the time fact-checking gets underway, people have already settled into their conclusions.”
