Facebook has struggled for months over whether it should crack down on false news stories and hoaxes that are being spread on its site. Now, it has finally come to a decision.
The social network is going to partner with the Poynter International Fact-Checking Network, which includes groups such as Snopes and the Associated Press, to evaluate articles flagged by Facebook users. If those articles do not pass the smell test for the fact-checkers, Facebook will label that evaluation whenever they are posted or shared, along with a link to the organization that debunked the story. Many of the organizations said that they’re not getting paid for this.
“We have a responsibility to reduce the spread of fake news on our platform,” Adam Mosseri, Facebook vice president of product development, told The Washington Post. Mosseri said the social network still wants to be a place where people with all kinds of opinions can express themselves but has no interest in being the arbiter of what’s true and what’s not for its 1 billion users.
The new system will work like this: If a story on Facebook is patently false — saying that a celebrity is dead when they are still alive, for example — then users will see a notice that the story has been disputed or debunked. People who try to share stories that have been found false will also see an alert before they post. Flagged stories will appear lower in the news feed than unflagged stories.
Users will also be able to report potentially false stories to Facebook or send messages directly to the person posting a questionable article.
The company is focusing, for now, on what Mosseri called the “bottom of the barrel” websites that are purposefully set up to deceive and spread fake news, as well as those that are impersonating other news organizations. “We are not looking to flag legitimate organizations,” Mosseri said. “We’re looking for pages posing as legitimate organizations.” Articles from legitimate sites that are controversial or even wrong should not get flagged, he said.
There is no blacklist of sites that will automatically be labeled, Mosseri said. But Facebook has built a sort of data profile of characteristics that fake-news articles share — such as low share numbers after the headline is clicked — which it will use to decide when to have something fact-checked.
The company will also prioritize checking stories that are getting lots of flags from users and are being shared widely, to go after the biggest targets possible.
If someone wants to appeal a label, they can direct that complaint to the fact-checking organization that made the call on whether an article was false.
It’s a little hard to evaluate how effective Facebook’s program might be, and whether the social network is doing enough to fight the fake news problem, said Duy Linh Tu, professor at the Columbia graduate school of journalism. “If I’m a technology company, it seems like this is above and beyond…if the goal is providing your audience with the best content possible,” he said. “From a journalistic side, is it enough? It’s a little late.”
Plus, he said, Facebook is fine to filter out other content — such as pornography — for which the definition is unclear. There’s no clear explanation for why Facebook hasn’t decided to apply similar filters to fake news. “I think that’s a little weak,” Tu said. “If you recognize that it’s bad and journalists at the AP say it’s bad, you shouldn’t have it on your site.”
Others said Facebook’s careful approach may be warranted. “I think we’ll have to wait and see early results to determine how effective the strategy is,” said Alexios Mantzarlis, of Poynter’s International Fact-Checking Network. “In my eyes, erring on the side of caution is not a bad idea with something so complicated,” he said.
Facebook is also trying to crack down on people who have made a business in fake news by tweaking the social network’s advertising practices. Any article that has been disputed, for example, cannot be used in an ad. Facebook is also playing around with ways to limit links from publishers with landing pages that are mostly ads — a common tactic for fake-news websites.
All of these efforts, he said, are works in progress. Users will start seeing them Thursday, but Facebook is testing different options to see what works best. This round of efforts is the first of many that Facebook will try.
“We don’t think it will get us all the way there,” he said. “I expect it to be something we need to invest in on an ongoing basis.”
Paul Horner, a fake news writer who makes a living writing viral hoaxes, said he wasn’t immediately worried about Facebook’s new crackdown on fake news sites. “It’s really easy to start a new site. I have 50 domain names. I have a dedicated server. I can start up a new site within 48 hours,” he said, shortly after Facebook announced its new anti-hoax programs. If his sites, which he describes as “satire”-focused, do end up getting hit too hard, Horner says he has “backup plans.”