DigitalMANIPULATIVE

Participatory Disinformation

What it is

Disinformation produced collaboratively — elites supply frames and cues, ordinary supporters generate the specific claims, sightings, and evidence, and the elites amplify what the crowd produced — so that a campaign has no single author and every participant is sincere.

How it works

Strategic information operations are usually pictured as top-down: a sponsor writes the lie and pays for its spread. Kate Starbird and colleagues (2019) showed that the most effective ones are collaborative work. Political leaders and influencers set expectations (the election will be rigged; the shooting looks staged), audiences primed by those expectations interpret ambiguous events accordingly and post their interpretations as evidence, and the leaders amplify the most vivid of those posts back to the crowd as confirmation. Starbird, DiResta, and DeButts (2023) traced this loop through the 2020 election-fraud narratives: a poll worker moving a box, a truck at a loading dock, or an overnight count update became fraud through the crowd's reading, not through fabrication at the top. The lever is sincerity. Participants believe what they post, which defeats bot detection and makes fact-checking feel like an attack on ordinary people; the loop also lets elites cite the groundswell as independent corroboration of what they seeded. The tell is a frame that precedes the evidence, and evidence that only ever confirms it.

Real-world examples

  • The Election Integrity Partnership's Long Fuse report (2021) documented how 2020 fraud claims moved from local posts (Sharpiegate in Arizona, the suitcases of ballots in Georgia) to influencers and campaign officials and back, with each cycle adding certainty and none adding verification.
  • After the 2013 Boston Marathon bombing, Reddit users crowdsourced an investigation that wrongly identified a missing student, Sunil Tripathi, as a suspect; the site apologized after his family was harassed, and his death was found to be unrelated to the attack.
  • From 2017, a loose online community around writers such as Louise Mensch and Claude Taylor produced a stream of claims about sealed indictments and imminent arrests in the Russia investigation; that year a hoaxer told the Guardian he had fed Taylor fabricated sources to see whether the claims would be checked, and they had not been.
  • QAnon is participatory by design: cryptic drops invited followers to do the research and bake the crumbs into narratives, and the movement's most consequential claims were authored by its audience rather than by Q.
  • The contrast shows the difference: Bellingcat's open-source investigations of MH17 and the Skripal poisoning were crowd-assisted but demanded geolocation, timestamps, and public correction, which participatory disinformation never supplies.

Ethical guidelines

  • Leaders who tell supporters what to expect, then amplify what supporters report, are authoring a campaign while disclaiming authorship of any claim in it.
  • Sincere belief does not make a false accusation harmless; participants who name real people as fraudsters or actors are responsible for the harm those people suffer.
  • Honest crowd investigation publishes its methods, invites refutation, and corrects publicly; anything that does not is not investigation.

How to defend against it

  • Watch for the loop: when a leader predicts an outcome, supporters then find it, and the leader cites the finding, the evidence was produced by the prediction rather than by the world.
  • Before sharing a citizen sighting, apply SIFT — stop, investigate the source, find better coverage, trace the claim to its original context — and notice when the original is an interpretation rather than an observation.
  • Ask what a sighting rules out: genuine evidence of fraud would look different from an ordinary process misread, and the poster can rarely say how.
  • If you take part in online investigation, adopt the disciplines that distinguish it from rumor — geolocate, timestamp, name what you could not verify, and post corrections as prominently as claims.
  • Resist treating the size of a sincere crowd as corroboration; a thousand people reading the same ambiguous clip through the same frame constitute one observation, not a thousand.

From the Defense Playbook

Every playbook entry states how strong its evidence is and when not to use it. Browse the full playbook.

References

  1. Starbird, K., Arif, A., & Wilson, T. (2019). Disinformation as Collaborative Work: Surfacing the Participatory Nature of Strategic Information Operations. Proceedings of the ACM on Human-Computer Interaction, 3(CSCW), Article 127
    The framework of information operations as collaborative work between orchestrated and organic participants.
  2. Starbird, K., DiResta, R., & DeButts, M. (2023). Influence and Improvisation: Participatory Disinformation during the 2020 US Election. Social Media + Society, 9(2)
    The top-down and bottom-up loop through which 2020 election-fraud narratives were produced.
  3. Election Integrity Partnership (2021). The Long Fuse: Misinformation and the 2020 Election. Stanford Internet Observatory, University of Washington Center for an Informed Public, Graphika, and DFRLab
    Case documentation of how local claims were amplified by influencers and officials and fed back to audiences.
Last reviewed
Suggest a correction

Detect Participatory Disinformation in any text

Paste any message, email, or article into our free Manipulation Detector to see if Participatory Disinformation or other techniques are being used on you.

Related Articles