Fact checking is a critical step in our process of verifying or debunking misinformation. This independent step involves calling sources, consulting public data and authenticating images and videos. Often, this happens at scale across many countries through a variety of mechanisms such as Meta’s content review technology that predicts if something is likely to be false or inaccurate in different contexts and communities; or Threads, an open group where people can flag posts they think should be reviewed. Fact checkers may also identify content to review on their own, based on the specifics of the local environment.
After receiving the research package and doing a top read, the fact checker creates a copy of the annotated draft without footnotes and removes them from the document—this is called the “checking copy.” The fact checker saves this copy in their (and their publication’s) fact-checking folder for reference later.
At this point, the fact checker begins to call every person quoted or cited by the reporter in the piece (on background) as well as any additional people that were interviewed but not mentioned in the story for verification purposes. When contacting sources, the fact checker makes clear that they are reaching out to them for verification and not for further reporting.
If a source is not available to be contacted by the fact checker, the fact checker may try to reach out to them again using alternate methods such as email or Twitter direct message, and they will explain that they are doing this in accordance with the Two-Layer Principle. If this isn’t possible, the fact checker consults with the head of research to discuss other options.