Facebook does share info about viral news articles with its fact-checking partners, but CAP suggests that Facebook’s own team take a look at especially viral content. That makes a lot of sense, and you might wonder why Facebook hasn’t had a virality circuit breaker all along.

The pandemic and misinformation around COVID-19 have highlighted Facebook’s struggle against bogus content. The platform has already removed seven million posts for coronavirus misinformation, including a post from President Trump’s account. Facebook labeled another 98 million posts as false, but not worthy of outright removal.

Facebook has tried to fight against COVID-19 “myths” with a “Facts About COVID-19” info center, but curbing its algorithms could be a more effective approach. Last year alone, health misinformation networks generated an estimated 3.8 billion views on Facebook, and a new report by Avaaz found that Facebook’s algorithms were largely to blame.

The company removes posts that spread false claims about cures or treatments for COVID-19, as well as other misinformation health organizations say is dangerous. The company has also taken steps to push credible health information, including debunking common rumors about the virus and pushing out PSAs about wearing masks. 

Despite these efforts, misinformation about the pandemic has been rampant on both Facebook and Instagram, and the company’s stricter policies have been repeatedly tested.

In May, a viral video that falsely claimed masks make people sick and that the coronavirus was created in a lab racked up millions of views before Facebook removed it. The scenario repeated itself last month when another video, which falsely claimed that the anti-malaria drug hydroxychloroquine was a “cure” for COVID-19, got more than 20 million views in a single day before Facebook took action against it. In both cases, copies of the videos continued to spread across Facebook and Instagram well after the removals began. 

Last week, Facebook removed a post from Donald Trump after he claimed that children are “almost immune” from COVID-19. That video was also up for several hours and was viewed millions of times before it was removed.

People who run popular pages with large followings are able to appeal fact-checking decisions. And when conservatives who have criticized the social network for being biased against them appeal these decisions, Facebook has reportedly flagged the issue as requiring “escalation.” It’s then that Facebook’s team, “with direct oversight from company leadership,” steps in.

“The list and descriptions of the escalations, leaked to NBC News, showed that Facebook employees in the misinformation escalations team, with direct oversight from company leadership, deleted strikes during the review process that were issued to some conservative partners for posting misinformation over the last six months,” NBC writes. “The discussions of the reviews showed that Facebook employees were worried that complaints about Facebook’s fact-checking could go public and fuel allegations that the social network was biased against conservatives.”

Pages for Breitbart, Donald Trump Jr., Eric Trump, the Gateway Pundit and Diamond and Silk have all gotten a pass under this process, according to NBC.

In a statement, a Facebook spokesperson said the company’s policy allows it to “determine that one or more of those ratings does not warrant additional consequences,” but didn’t elaborate on the process.

“We defer to third-party fact-checkers on the rating that a piece of content receives. When a fact checker applies a rating, we apply a label and demotion,” the spokesperson said. “But we are responsible for how we manage our internal systems for repeat offenders. We apply additional system wide penalties for multiple false ratings, including demonetization and the inability to advertise, unless we determine that one or more of those ratings does not warrant additional consequences. To this day, we remain the only company that partners with over 70 fact-checking organizations to apply fact-checks to millions of pieces of content.”

The report from NBC comes one day after BuzzFeed News also found that Facebook officials, including  Brett Kavanaugh bff and global policy head Joel Kaplan, had intervened in order to help prominent conservatives get fact checks removed or otherwise avoid consequences for sharing misinformation. The social network later fired an employee who questioned these decisions on internal forums, according to BuzzFeed.

Twitter knows its system for automatically fact-checking posts mentioning 5G and the coronavirus is more than a little overzealous, and it’s hoping to tone things down. The social network has promised to “improve” the precision of its labeling system and fact-check fewer tweets that merely happen to include certain keywords. The company is developing “new automated capabilities” to focus more on genuinely relevant tweets, according to an additional statement.

There was no indication of a timeline for the improvements.

Google’s fact checking now extends beyond standard searches and YouTube. The internet pioneer has introduced fact check labels for image search results. Tap a bogus picture for a detailed view and you’ll see a blurb from a verified source indicating what’s false and offering a link to the full article debunking the image. If a photo is a known fake, you’ll find out before you start sharing it with your friends in disbelief.

The company is using the open ClaimReview method (also used for YouTube) to provide fact checks from a variety of “independent, authoritative” sources like PolitiFact.

Twitter told Engadget that its labels are an “iterative process” and that it was focusing on tweets mentioning 5G and COVID-19 for now. The company also acknowledged that it might make mistakes, and that it would refine its process over time. You can read the official statement below.

The fact checks could get irksome with this kind of ubiquity, and might desensitize the people who most need to see them. They’re certainly not very bold. At the same time, this at least shows that Twitter is using checks more freely, and might catch misinformation peddlers who otherwise slip through the cracks.

“Labeling or placing a warning on Tweets continues to be an iterative process.  Given the global spread of misinformation and disputed claims around 5G and COVID-19, we prioritized labeling tweets with that information. Our team is currently reviewing other types of content and will label additional tweets soon.”

Twitter CEO Jack Dorsey fired back in a series of tweets. “We’ll continue to point out incorrect or disputed information about elections globally. And we will admit to and own any mistakes we make. This does not make us an ‘arbiter of truth.’ Our intention is to connect the dots of conflicting statements and show the information in dispute so people can judge for themselves. More transparency from us is critical so folks can clearly see the why behind our actions.”

This does not make us an “arbiter of truth.” Our intention is to connect the dots of conflicting statements and show the information in dispute so people can judge for themselves. More transparency from us is critical so folks can clearly see the why behind our actions.

— jack (@jack) May 28, 2020

Twitter subsequently labeled hundreds more tweets with fact-checking labels, the New York Times reported.

In the same interview with Fox News, Zuckerberg said he believes Facebook has a “stronger” record on defending free speech than other companies, but didn’t provide evidence of that claim. Facebook works with dozens of fact-checking organizations, but doesn’t allow posts from politicians to be fact checked — a policy that has been widely criticized.

Twitter didn’t place limits on Trump’s tweets about election fraud or make them less visible. Instead the company added a label, telling users to “get the facts” about mail-in ballots, and linked to publications that had fact-checked the claims. A Twitter spokesperson said the labels were added because the comments were “potentially misleading.”

Zuckerberg’s criticism comes as Trump signed an executive order that will attempt to put new restrictions on how social media platforms moderate content. It’s unclear how the order will be enforced, or how much impact it would ultimately have. 

The Facebook CEO, who had a private dinner with the president last year, said he doesn’t believe it’s the “right reflex” to place limits on social media companies. “In general, I think a government choosing to censor a platform because they’re worried about censorship doesn’t exactly strike me as the right reflex there,” he said.

Then, in an interview with Fox News Wednesday morning, Kellyanne Conway singled out one Twitter employee by name: Yoel Roth, the company’s head of site integrity. Referring to him as the “head of integrity,” Conway spelled out his Twitter handle and said “somebody in San Francisco go wake him up and tell him he’s about to get more followers.”

On Fox & Friends, Kellyanne Conway appears to direct online harassment at Twitter’s head of site integrity, Yoel Roth: “Somebody in San Francisco will wake him up and tell him he’s about to get a lot more followers.” pic.twitter.com/H9ceUu6Ezv

— Bobby Lewis (@revrrlewis) May 27, 2020

Conway also claimed that Roth is “constantly attacking Trump voters,” an apparent reference to years-old tweets criticizing the Trump Administration. Fox News also seized on the comments and stated that Roth was “in charge” of Twitter’s decision to fact-check the president. Trump’s sons and Trump’s official campaign account also shared tweets singling out Roth to their millions of followers.

Twitter confirmed that Roth was not singlehandedly responsible for the decision to fact check Trump. Roth is part of Twitter’s larger trust and safety team, which makes policy decisions. As others have pointed out, Roth primarily deals with platform abuse  — misuse of the company’s API, spam and bot campaigns — not fact checking or content decisions.

“No one person at Twitter is responsible for our policies or enforcement actions, and it’s unfortunate to see individual employees targeted for company decisions,” a Twitter spokesperson said in a statement. 

But Twitter’s comments have done little to quiet the angry mob of Trump supporters, many of whom have long accused the company of being biased against conservatives. Over the last day, Roth has been inundated with harassment and even death threats, Protocol reported.

For Twitter, the incident is yet another reminder of how much work it still needs to do if it hopes to fix its harassment problem. For all of the new reply-limiting controls and anti-bullying tools, there’s little any one user can do when thousands of users swarm one person all at once.

Twitter

Twitter has appended a link urging users to “get the facts” about mail-in ballots. Clicking through, Twitter further notes that fact checkers say there is no evidence to support these claims and that voting by mail is already used in a number of states. The company also links to stories from CNN, the Washington Post and other outlets debunking these claims.

The labels mark the first time the company has fact-checked the president, a move Facebook and other companies have largely resisted until now. The fact check also comes after Twitter came under fire for declining to remove a series of tweets from Trump promoting a debunked conspiracy theory about the death of a former staffer of MSNBC host Joe Scarborough.

Twitter

In a statement to Engadget, a Twitter spokesperson said the tweets were labeled due to “potentially misleading information,” but declined to say if the company planned to label more of the president’s tweets. Twitter has previously added a “manipulated media” label to a deceptively-edited video Trump retweeted.

“The Tweets contain potentially misleading information about voting processes and have been labeled to provide additional context around mail-in ballots. This decision is in line with the approach we shared earlier this month,” a Twitter spokesperson said. 

Update 5/26 8pm ET: In a series of follow-up tweets, Trump responded to Twitter’s fact-checking saying the company was “stifling” free speech.

….Twitter is completely stifling FREE SPEECH, and I, as President, will not allow it to happen!

— Donald J. Trump (@realDonaldTrump) May 26, 2020

YouTube is widening the audience for its fact check panels, and not a moment too soon. The video service now displays (via TechCrunch) independent fact checks for searches of known false claims. Look for claims behind COVID-19 conspiracy theories, for instance, and a panel will link you to an article debunking those myths. As before, YouTube is limiting these cards to relatively narrow searches — you shouldn’t get them while you’re simply looking for the latest news.

The panels bring up info from third-party publishers who honor the International Fact-Checking Network’s ethics code and meet the standards for ClaimReview’s tagging system. This includes partners like FactCheck.org, PolitiFact and The Washington Post Fact Checker.