Statement from NewsGuard co-CEOs Steven Brill and Gordon Crovitz on Facebook’s Promise to Remove COVID-19 Vaccine Misinformation from its Platform
(December 3 — New York City) It’s encouraging to see Facebook talking about the important problem of COVID-19 vaccine misinformation—and we hope that they will now do more than just talk.
Early in the COVID-19 pandemic, Facebook put out a similar statement saying it was removing COVID-19 misinformation from its platform—but when our analysts checked, we found ‘super-spreader’ accounts reaching tens of millions of users pushing misinformation about the pandemic, including false, harmful medical advice. And only 37% of the false posts we checked had any kind of warning attached to them. In fact, as we documented in reports we provide to the World Health Organization, Facebook not only failed to counter this misinformation effectively, in many cases, it also was actively recommending the misinformation to users.
Facebook now says it will take down COVID-19 vaccine misinformation, and this time, we hope they will take a different approach. In the past, Facebook has relied on a mix of black-box artificial intelligence algorithms with no transparency and anonymous moderators. The lack of disclosure and transparency allows conspiracy theorists, including those peddling these hoaxes, to double down on existing conspiracies and create new conspiracies about cover-ups. And because Facebook’s attempts to counter misinformation are at the level of individual posts, they are forced to fact-check billions of pieces of content, ensuring their mix of undisclosed humans and black box machine moderators will never be able to keep up.
At NewsGuard, we know that a huge percentage of misinformation comes from the same sources that repeatedly publish false information—indeed, 80% of the sources we found publishing COVID-19 misinformation had previously published other health misinformation and conspiracy theories. That’s why we provide transparent, accountable, and constantly updated ratings of news and information sources that platforms can use to understand which sources need to be monitored or sanctioned, and which sources are trustworthy enough to promote algorithmically. By monitoring these sources, it is possible to identify the most important myths and misinformation narratives quickly for use with AI tools—which we offer through our newly developed Misinformation Fingerprints™ catalogue of top myths and misinformation narratives. These narratives, with transparently sourced debunkings, include numerous hoaxes about the COVID-19 vaccine and other topics and are designed for use by AI tools, including Facebook’s.
We will continue to monitor COVID-19 vaccine myths and the accounts and publishers spreading them and report that information to the WHO, other interested parties, and the public. We are glad to be working with other platforms and technology companies to buttress their work combatting misinformation.
- Steven Brill, Co-CEO, firstname.lastname@example.org, +1 212-332-6301
- Gordon Crovitz, Co-CEO, email@example.com, +1 212-332-6407