Disinformation at machine speed: clustering stories across channels
When the same framing appears on nine channels in ninety minutes, is it public reaction or a coordinated push? Story clustering is how you tell the difference.
Coordinated influence operations exploit a simple asymmetry: they can push a narrative across many channels faster than any human team can watch them. By the time an analyst notices a pattern on one screen, the same framing has already propagated across a dozen.
Story clustering closes that gap. CLERINT Media groups segments that share keyword patterns across two or more streams into a single emerging story — automatically, as it happens. Instead of nine isolated broadcasts, you see one narrative with nine surfaces.
That grouping is what makes coordination visible. A story that surfaces organically tends to spread unevenly and attribute to many origins. A seeded one propagates from a small number of source streams outward, on a suspiciously tight timeline. The cluster's shape tells you which you are looking at.
None of this is a verdict on its own — clustering surfaces a pattern for a human to judge, with the underlying segments retained as evidence. But it turns a question that used to take days of manual review into one an analyst can answer while the campaign is still live.