technology

False attacks on Facebook could bring 'a Titanic-sized disaster' in 2020

Misinformation is still widely accessible on the world’s largest social network.

Mark Zuckerberg

False news reports that attack U.S. politicians have been viewed more than 150 million times on Facebook since the beginning of 2019, according to an analysis published Wednesday that points to a growing threat of deception swamping next year's elections.

The analysis, from the activist group Avaaz, highlights how misinformation — often targeting political figures like House Speaker Nancy Pelosi and President Donald Trump — is still widely accessible on the world’s largest social network despite pledges by Mark Zuckerberg, the company's chief executive, to combat misinformation and other false reports on the global network.

The study also notes a growing sophistication in how people are being targeted on Facebook: Political actors are moving away from a strategy that focuses solely on bombarding individuals with paid-for partisan messages, toward more complex tactics that rely on presenting would-be voters with traditional shared content that does not face the same scrutiny as political ads.

Avaaz could not say if the sharing of these false reports was part of a wider online deception campaign or merely Facebook users sharing material that aligned with their political beliefs.

“We’re heading towards a Titanic-sized disaster,” Fadi Quran, campaign director at Avaaz, told POLITICO in reference to the 2020 election. “Facebook is not being transparent about the problem. There’s no telling how much disinformation content is being spread.”

The findings come as Facebook continues to face criticism, especially from Democrats, for its policy of refusing to fact check campaign ads and its failure to take down some kinds of false political attacks. Those include an ad by Trump's reelection campaign making baseless accusations of corruption against former Vice President Joe Biden, as well a doctored video last spring that falsely portrayed Pelosi as drunkenly slurring her words.

In response to the study, Facebook said that people continued to see false reports on its network, but said it had created digital warnings labels that were tagged to content that could be potentially false. Much of the content that Avaaz had flagged, though, did not have a warning label.

As part of the analysis, the group looked at the 100 political fake news stories that had been most frequently commented on or shared on Facebook between Jan. 1 and Oct. 31, using content that had already been debunked by independent third-party fact-checkers. The group based its figures on publicly available Facebook data and third-party social media tools to rank which posts had been viewed the most.

In total, the false material — mostly stemming from right- or left-wing partisan websites that claimed to be legitimate media outlets — had been posted 2.3 million times across Facebook and had garnered almost 9 million interactions through either comments, "likes," or sharing to other Facebook pages. The group estimated that, in total, the misinformation had been viewed 158 million times since the beginning of the year.

Almost 40 percent of the debunked material was spread by individual Facebook users, a further 35 percent was shared by non-official political actors and 19 percent was disseminated by alternative media Facebook pages. Only 1 percent of the content was spread by traditional media outlets, according to Avaaz.

Almost two-thirds of the material targeted either Democratic politicians and left-wing causes, according to the analysis. That included false claims that Pelosi had diverted $2.4 billion from Social Security spending to pay for the impeachment hearings into Trump. Pelosi, former President Barack Obama and New York Democratic Rep. Alexandria Ocasio-Cortez were some of the politicians most targeted by the fake reports, Avaaz said.

Roughly the other third of false news reports focused on Republican lawmakers and right-wing issues, with Trump by the far the largest target. Those included fake allegations that his grandfather was a member of the Ku Klux Klan.

As political campaigners gear up for next year’s vote, Avaaz also noted that the level of Facebook interactions with these false reports had significantly increased during the last three months as political messages began to be shared more virally among potential voters.

Quran, the group’s campaign director, said this was partly due to the evolving tactics of misinformation, which included digital tricksters latching on to existing content on Facebook to spread their political messages. Many of these tactics rely on a steady flow of false news reports from websites outside Facebook’s digital ecosystem that are then shared widely by the company’s users.

Such content-sharing allows untrue messages, often about lawmakers and political causes, to be shared with little, if any scrutiny. Despite the false claims in these posts, Facebook has routinely refused to remove the material either on free speech grounds or because the content does not breach its terms of service.

Following the 2016 election, Zuckerberg said his company would clamp down on such misinformation through a combination of third-party fact-checkers and new technology that would siphon out the worst offenders. But with 12 months to go before the U.S. again heads to the polls, it’s unclear how successful the social network has been in removing such debunked content from its platform.

“This is disinformation 2.0,” Quran said. “Malicious actors are now looking for existing content and amplifying that for their own purposes. It’s hard to detect that.”