Facebook is partnering with news wire service Reuters to fact-check false stories on Facebook, the social network announced today. The two will launch a pilot “Wide Scale Reporting Program” with the goal of “expanding it to other news organizations and communities around the world.” This is an extension of an existing partnership between Facebook and Reuters on fact-checking in Ireland, and it could signal Facebook’s awareness that it can’t rely on its own algorithms and its staff to hold content accountable for falsehoods.
Reuters will review Facebook photos, videos, and text to verify its authenticity before becoming the first outlet to select a set of selected fakes to go through Facebook’s fact-checking process. Once properly confirmed, Facebook will either display the verified profile of the given Facebook profile along with the Reuters tag or confirm its status as false with a yellow box with a red flag. Reuters will then publish the announcement of its fact-checking to its own social media platforms.
The pilot gives verified content its own feed, and other outlets can take or apply for the contact information of the Reuters fact-checker, should they have concerns about the quality of a given post. Reuters may also receive freelance work from its network of over 200 reporters around the world, though it wouldn’t share the exact amount or scope of the program.
The partnership’s first test was Ireland, where Reuters confirmed around 750 articles from deepfakes. But the pilot also includes two additional countries, which Facebook says it will disclose later, and it has a long list of future candidate countries. Right now Reuters simply has to check its platform for Reuters tags before publishing. But it does save Facebook time while the Wall Street Journal, CNN, BuzzFeed, Forbes, NBC, and a bevy of start-ups have already launched fact-checking campaigns of their own.
Facebook has admitted that its algorithms do not always make sense. But relying on formulas to support moderation means that it too often couldn’t detect Russian posts using inflammatory topics like Black Lives Matter to fuel the 2016 U.S. election. Meanwhile, if Facebook’s humans are easily fooled by deepfakes and other contrived objects, the social network will then need to add more moderators and technology to keep up. That could make it more difficult for Facebook to scale as it faces an ever-increasing number of fake accounts and dangerous ads. The need for extra screening is likely on Mark Zuckerberg’s mind as he is stepping down as CEO.
As a response to criticism, Facebook beefed up its photo verification tools. Now if you upload a photo you think could be fake, Facebook will deliver an image-matching feature to verify. It also made it easier to flag questionable items that you can place on a flag list. And Facebook recently announced a partnership with PolitiFact to identify false political claims.
By partnering with Reuters and raising its profile through its social media platforms, Facebook can gain better credibility and credibility could then trickle down to boost its Trust Graph. But verification from Reuters will have its own challenges. It will have to read dozens of obfuscating websites, often years out of date. Moreover, from its own coverage, its top team reports to chief news executive Jeff Marsilio. It also probably won’t agree with Reuters editorial standards, which have shifted significantly over the years. That means plenty of misinformation could slip through its blinders.