Skip to main contentSkip to navigationSkip to navigation
Facebook has enlisted fact-checking partners, but debunked articles frequently remain on the site without a warning.
Facebook has enlisted fact-checking partners, but debunked articles frequently remain on the site without a warning. Photograph: Niall Carson/PA
Facebook has enlisted fact-checking partners, but debunked articles frequently remain on the site without a warning. Photograph: Niall Carson/PA

Facebook promised to tackle fake news. But the evidence shows it's not working

This article is more than 6 years old

Following pressure from users, the social network introduced tools to stem the spread of false information. But the rollout has been rocky at best

When Facebook’s new fact-checking system labeled a Newport Buzz article as possible “fake news”, warning users against sharing it, something unexpected happened. Traffic to the story skyrocketed, according to Christian Winthrop, editor of the local Rhode Island website.

“A bunch of conservative groups grabbed this and said, ‘Hey, they are trying to silence this blog – share, share share,’” said Winthrop, who published the story that falsely claimed hundreds of thousands of Irish people were brought to the US as slaves. “With Facebook trying to throttle it and say, ‘Don’t share it,’ it actually had the opposite effect.”

The spreading of Winthrop’s piece after it was debunked and branded “disputed” is one of many examples of the pitfalls of Facebook’s much-discussed initiatives to thwart misinformation on the social network by partnering with third-party fact-checkers and publicly flagging fake news. A Guardian review of false news articles and interviews with fact-checkers and writers who produce fake content suggests that Facebook’s highly promoted initiatives are regularly ineffective, and in some cases appear to be having minimal impact.

Articles formally debunked by Facebook’s fact-checking partners – including the Associated Press, Snopes, ABC News and PolitiFact – frequently remain on the site without the “disputed” tag warning users about the content. And when fake news stories do get branded as potentially false, the label often comes after the story has already gone viral and the damage has been done. Even in those cases, it’s unclear to what extent the flag actually limits the spread of propaganda.

The warning message that appears when some Facebook users try to post a fake news article. Photograph: Facebook

The social network’s efforts to curb fake news followed widespread backlash about the site’s role in proliferating misinformation during the 2016 presidential election. The rocky rollout of Facebook fact-checking is as much a product of the enormity of the problem of internet propaganda as it is a reflection of what critics say is a failure by the company to take this challenge seriously.

“Fake news is flying thick and fast,” said Brooke Binkowski, the managing editor of Snopes, one of the popular fact-checking sites partnering with Facebook to review content that users flag as false. Asked if she believes Facebook’s new system is having an effect on false news through the use of tags, she added: “I honestly can’t tell.”

Last year, Facebook faced growing criticisms that it may have helped Donald Trump get elected by allowing fake election news to outperform real news, and creating filter bubbles that facilitated the increasing polarization of voters. In response, Facebook announced that it would work to stop misinformation in part by letting users report fake news articles, which independent fact-checking groups could then review.

When two or more fact-checkers debunk an article, it is supposed to get a “disputed” tag that warns users before they share the piece and is attached to the article in news feeds, a feature rolled out in March.

While some of the fact-checking groups said the collaboration has been a productive step in the right direction, a review of content suggests that the labor going into the checks may have little consequences.

ABC News, for example, has a total of 12 stories on its site that its reporters have debunked as part of its Facebook partnership. But with more than half of those stories, versions can still be shared on Facebook without the disputed tag, even though they were proven false. That includes several of the original stories that ABC bunked, such as an AngryPatriotMovement.com article about Obama planning a coup, an EmpireNews.net story that Obama built a statue of himself for the White House, and a USANewsToday.org article about Republicans taking secret payments from Hillary Clinton.

Xana O’Neill, the managing editor of ABC News Digital, said her group has done roughly two-dozen Facebook fact checks and said she believed the process was having a positive impact. But, she added: “It’s hard to know the scope of it without the Facebook numbers.”

Facebook refused to provide data or information on the number of articles that have been tagged as disputed, how a flag impacts traffic and engagement, if there are specific websites most frequently cited and how long after publication the flags are typically added. A spokesman said “we have seen that a disputed flag does lead to a decrease in traffic and shares”, but declined to elaborate.

Robert Shooltz, who runs the site RealNewsRightNow, said it did not seem that the debunking of his content by Facebook fact-checkers was impacting traffic. Both Snopes and PolitiFact recently debunked his article that said Trump wants to bring back the military draft.

But the piece, which was published last year and recently began spreading again, does not have a disputed tag on it. It has had more than 33,000 likes, shares and comments, according to Facebook data.

“It’s had absolutely no effect. I’m happy about that,” said Shooltz, who considers his work satire, much like the Onion, and does not believe it should be targeted by Facebook fact-checking.

Facebook has faced criticism that it may have helped Trump get elected by allowing fake news to outperform real news on the platform. Photograph: UPI / Barcroft Images

Another piece of his, headlined “Pope Francis: God Has Instructed Me to Revise the Ten Commandments” was also debunked by Snopes and the AP in March and does have the “disputed” tag. But he said it wasn’t clear to him if there was any serious impact on traffic and noted that the article had already experienced a spike in views prior to the debunks.

Paul Horner, another well-known fake news writer, said some of his websites have been blocked on Facebook, but that other articles have gone unchecked, including one saying Trump issued an order allowing bald eagles to be hunted and another about the president canceling Saturday Night Live. Both were published on the St George Gazette fake news site and both were debunked by Snopes.

Chris Kitze, who runs BeforeItsNews.com, said that although he allows users to post any content without fact-checking, he hasn’t noticed Facebook tagging any of his site’s articles as fake news. That includes a recent piece debunked by Snopes claiming to include leaked photos showing how Obama practiced Islam in the White House.

“A lot of people think Obama is Muslim. That’s what it plays on. Is it real? I don’t know,” he said. “The fact is a lot of people thought it was real or it reflects their sentiment.”

Aaron Sharockman, the executive director of PolitiFact, the Pulitzer Prize-winning site that is also fact-checking for Facebook, said that by the time his reporters debunk an article, it could have been up for several days to a week or more, meaning the effect of the flag may be limited.

“We don’t have a great sense of the impact we’re having,” he said, adding “We haven’t seen anything from Facebook.”

Jestin Coler, a writer who got widespread attention for the fake news he published last year, said it was hard to imagine Facebook’s effort having any impact.

“These stories are like flash grenades. They go off and explode for a day,” said Coler, who said he is no longer publishing false news. “If you’re three days late on a fact check, you already missed the boat.”

He also noted that many consumers of fake news won’t be swayed by a “disputed” tag given their distrust of the media and fact-checkers: “A far-right individual who sees it’s been disputed by Snopes, that adds fuel to the fire and entrenches them more in their belief.”

A writer with TheLastLineofDefense.org, another site that posts fake news, said the website has “definitely seen a drop in traffic since Facebook started relying on outside fact-checkers”, but did not respond to further questions.

Melissa Zimdars, assistant professor of communication and media at Merrimack College, who created a viral list of untrustworthy news sites, said it seemed Facebook was largely responding to bad press: “My initial read on it is it’s ultimately kind of a PR move. It’s cheap to do. It’s easy. It doesn’t actually require them to do anything.”

A Facebook spokesperson said the fact-checking tags were just one tool in its ongoing efforts, which include taking action against fake accounts, disrupting financial incentives of fake news creators and launching an educational tool.

“We take seriously the issue of fighting false news and are utilizing an all-of-the-above approach. There’s no silver bullet solution, which is why we’ve deployed a diverse, concerted and strategic plan.”

More on this story

More on this story

  • Martin Lewis calls for publishers to act over fake news ads

  • How social media filter bubbles and algorithms influence the election

  • In Europe political attitudes to Facebook are changing

  • Facebook's only Dutch factchecker quits over political ad exemption

  • Will fake news wreck the coming general election?

  • Stephen Fry: Facebook and other platforms should be classed as publishers

  • The Audio Long Read
    Why can’t we agree on what’s true any more? – podcast

  • 'Chilling': Singapore's 'fake news’ law comes into effect

  • Facebook fined £94m for 'misleading' EU over WhatsApp takeover

  • 2016: the year Facebook became the bad guy

Most viewed

Most viewed