Tech companies and election officials spent the past three years working to address the vulnerabilities that allowed Russia to promote disinformation and sow political divisions ahead of the 2016 election.
On Monday night, though, it didn't take foreign interference to highlight the persistent vulnerabilities around the 2020 election. Americans proved perfectly capable of spreading disinformation on their own.
"This might be a great wake-up call," said NBC cybersecurity analyst, Alex Stamos, director of the Stanford Internet Observatory and a former head of security at Facebook. "If it turns out that this disaster has very little long-term effect on the primary, it's a great demonstration of what could go wrong on election night."
One day after Americans cast their first votes in the election cycle, conspiracy theories, delayed election results and political opportunism intermingled online to create what disinformation researchers warn could be a harbinger of a self-inflicted worst-case scenario for the 2020 election.
Various pieces of false or misleading news circulated, including accusations of voter fraud and false claims that a major candidate had dropped out. Those claims were amplified by problems with the Iowa Democratic Party's reporting of the results, which were delayed in part because of problems with its smartphone app — which then generated its own conspiracy theories.
There is no indication that Russia or any other government or company pushed disinformation around the Iowa caucuses.
"As we saw last night, there are domestic actors that are willing to do that and mysterious internet trolls, which could be domestic or foreign," Stamos said. "But even members of the media, who might be very strongly attached to one side or the other, are happy to amplify."
The most viral piece of disinformation came from the conservative legal group Judicial Watch, which falsely claimed that eight Iowa counties had more voter registrations than their numbers of citizens. Despite efforts by Paul Pate, Iowa's secretary of state, to debunk the information by pointing to public county-by-county voter registration totals, the claims were repeated by major conservative media outlets, including The Epoch Times.
Facebook told NBC News that the company's elections operations center was up and running and that it had been communicating with Pate, the Democratic National Committee and third-party fact-checkers throughout the day. Facebook eventually put warnings on several posts that repeated the Judicial Watch claim indicating that they contained false information.
But that came hours after the story began to spread. The Epoch Times post garnered 175,000 Facebook comments, likes and shares, according to the social media analysis tool Buzzsumo.
On Twitter, where the false claim first gained traction, a company representative said the claims did not violate the company's election integrity policy, "as it does not suppress voter turnout or mislead people about when, where or how to vote."
The specificity of both platforms' rules around disinformation leave room for known disinformation agents to work during elections, said Joan Donovan, director of the Technology and Social Change Research Project at Harvard University's Shorenstein Center on Media, Politics and Public Policy.
The voter fraud claim was a "classic disinformation tactic," she said, known as "trading up the chain," where false stories are repackaged by blogs and shared.
Donovan said that while the specific intent of the disinformation — to disrupt the caucuses or sow larger distrust in the national viewing audience — may have been unclear, the goal of attaining mainstream media attention was met.
"The fact that a known disinformation operation, like The Epoch Times, enjoyed hyper-engagement on Facebook shows that Facebook has not been able to stop the weaponization of their product by those with some well-placed resources," Donovan said.
The Judicial Watch claim was far from the only piece of misleading news to circulate around the Iowa caucuses. Just days earlier, conspiracy theories were pushed by supporters of Sen. Bernie Sanders, I-Vt., and entrepreneur Andrew Yang over the canceled Des Moines Register poll.
News that the results of the caucuses would be delayed because of inconsistencies in data reporting sparked a new wave of conspiracies — most notably from surrogates for President Donald Trump and campaign officials who claimed the delay was evidence of some sort of vote tampering by Democrats.
"Quality control = rigged?" Trump's campaign manager, Brad Parscale, tweeted.
The president's adult sons echoed Parscale's claims in "rigged" tweets of their own.
Some Democrats also seized on another undocumented conspiracy theory, suggesting that former South Bend, Indiana, Mayor Pete Buttigieg and his campaign had worked with the maker of the troubled Iowa caucuses app to tilt the results in his favor, tweeting on the "#MayorCheat" hashtag.
Sen. Mark Warner, D-Va., said in a statement that the caucuses showed that the election system remains at risk.
"What happened in Iowa last night underscores the necessity of all these measures were election-night systems to face a devastating hack," Warner said. "But what we've also seen [is] that this chaos has created an environment where misinformation is now running rampant online, further undermining confidence in the democratic process. As we've seen in the past, foreign actors like Russia and China won't hesitate to latch onto this kind of content in order to add to the domestic discord and distrust in our elections."
Not all disinformation attempts were so sophisticated. Right- and left-wing trolls got creative, repurposing a 2008 tweet from the NBC News Breaking News Twitter account to make it seem like Joe Biden was dropping out of the race.
Though many were joking, the tweet was pushed by well-known Twitter trolls who appear during breaking news events to spread false information.
NBC News appended an editor's note in the replies that clarified that the tweet was old, but still, it gained over 2,000 new retweets. Several users replied, "Got me."
