Facebook to stop using “disputed” labels on false news, will show fact checks as Related Articles after tests proved the latter method resulted in fewer shares
Scrolling through News Feed, it can be hard to judge what stories are true or false.
Context & Ripple Effects
Facebook's 'disputed' label was always a blunt instrument: the company took a full week to flag one made-up story when it began applying the labels in early 2017, and the badge only went on stories already rated false by fact-checkers. Through the spring and summer it quietly built the replacement — first testing links to fact-checks beside stories inside a Related Articles widget in April, then rolling out Related Articles for popular stories including flagged ones that August.
First-order effects
- Fact-checking partners' red 'disputed' badges disappear from the News Feed; their work now surfaces one click away, as articles listed under a story rather than stamped on it.
- The change is driven by Facebook's own test data — stories accompanied by fact-check articles were shared less than stories wearing a label, making this an engagement-measured retreat from visible moderation.
Second-order effects
- Once fact-checks live behind a click, the next levers are algorithmic: by the following spring Facebook was shrinking the size of fact-checked false stories in the Feed and using machine learning to prioritize what goes to fact-checkers at all [[a:929022]].
Third-order effects
- The arc runs from human-applied labels to machine-scaled suppression — detection, demotion, and contextual fact-checks layered invisibly into ranking — which makes platform moderation harder for users to see and audit even as it scales.
The trend: Content moderation on social platforms is shifting from visible, human-applied flags to invisible algorithmic demotion and contextual nudges, with share-rate data deciding which interventions survive.