When content has been rated by fact-checkers, we take action to (1) label it and (2) ensure fewer people see it, and (3) penalise repeat offenders. Meta's technology is designed to detect content that is the same or nearly identical to content rated by fact-checkers, applying notices and reduced distribution automatically. This integration operates across all content formats relevant to the service, including public posts, ads, articles, photos, videos, Reels, and text-only posts on both Facebook and Instagram.
When content has been rated by fact-checkers, we add a notice to it so people can read additional context. Content rated Satire or True won’t be labelled but a fact-check article will be appended to the post on Facebook. We also notify people before they try to share this content or if they shared it in the past. We use our technology to detect content that is the same or almost exactly the same as that rated by fact-checkers, and add notices to that content as well.
Ensuring fewer people see misinformation. Once a fact-checker has rated a piece of content as False, Altered or Partly False, or we detect it as near identical, it will appear lower in Feed on Facebook. We dramatically reduce the distribution of False and Altered posts, and reduce the distribution of Partly false to a lesser extent. Meta does not suggest content to users once it is rated by a fact-checker, which significantly reduces the number of people who see it.
Repeat offenders. Facebook pages, groups, profiles, and websites that repeatedly share content rated False or Altered will be put under some restrictions for a given time period. This includes removing them from the recommendations we show people, reducing their distribution, removing their ability to monetise and advertise, and removing their ability to register as a news Page.
Detection. Meta's systems support fact-checkers’ work through a signals-based detection approach, which uses various inputs - including user flags reporting “false information” - to identify and enqueue content for fact-checker review. Fact-checkers ultimately decide what to review and rate. Once content is rated, Meta applies automated enforcement actions (labelling, reduced distribution, ad rejection) and extends these actions to near-identical content detected through matching technology.
In terms of AI-generated content, fact-checkers may rate AI-generated media under our fact-checking programme policies. They often rely on AI experts and visual techniques to aid in the detection of this content.
