Earlier this month, the Post had obtained a hard drive belonging to the son of the current Democratic presidential nominee Joe Biden, and cited emails they found as indication of troubling business deals in Ukraine. Twitter responded by locking their account, but Facebook said it would “temporarily reduce distribution” of it.
According to Andy Stone, Facebook’s policy communications director, the story would have to be looked at by fact-checkers, which was part of a “standard process” to reduce the spread of “misinformation.”
On Friday, however, a Facebook statement seemed to confirm that no such review actually took place.
“As our CEO Mark Zuckerberg testified to Congress earlier this week, we have been on heightened alert because of FBI intelligence about the potential for hack and leak operations meant to spread misinformation,” a company spokesman said in a statement to the Guardian. “Based on that risk, and in line with our existing policies and procedures, we made the decision to temporarily limit the content’s distribution while our fact-checkers had a chance to review it.”
When that didn’t happen, we lifted the demotion.
The spokesman did not clarify whether the fact-checkers even attempted to verify the materials published by the Post, but no results of such a review have ever been published.
Moreover, the Guardian cited internal moderation documents that someone at Facebook leaked to them, indicating that the shadowban wasn’t part of a regular process. Instead, the documents purportedly showed the existence of a policy under which stories can be “manually enqueued” for suppression, citing the upcoming US elections as an “issue of importance” that justified such actions.
Who elected these people guardians of what is true?Facebook leak reveals policies on restricting New York Post's Biden story | Technology | The Guardian https://t.co/Tbng1r4Cdf
— CalThomas (@CalThomas) October 30, 2020
The documents also say that the standard practice is for Facebook AI to predict which content might contain misinformation, based on signals “including feedback from the community and disbelief comments,” in a sort of pre-crime enforcement straight out of dystopian science fiction.
There is also a de-facto “whitelist” of some 5,000 media outlets, called the “Alexa 5K,” whose content the AI ignores by default “under the assumption these are unlikely to be spreading misinformation.” The Post was apparently on that list, so the decision to suppress the story was made by actual Facebook employees, not an AI system flagging potential misinformation.
It worked, too. Data compiled by research firm NewsWhip and published in Newsweek last week showed that Facebook’s soft suppression was far more effective than Twitter’s outright censorship of the Post’s account, which was finally overturned on Friday.
On Facebook, the story reached only half as many readers as similar anti-Trump bombshells pushed by the mainstream media. Neither the claims by the New York Times about the president’s taxes nor the Atlantic’s repeatedly-refuted story about Trump calling US troops “suckers and losers” were flagged for fact-checking by either human moderators or Facebook’s AI.
Stone, Facebook’s policy director who announced the shadowban, is open about his history with the Democrats. His Twitter biography shows him working in the past for the Democratic Congressional Campaign Committee, Senator Barbara Boxer and Rep. Jerry McNerney, among others.
Like this story? Share it with a friend!