Science flags paper that found AI chatbots help debunk conspiracy theories
retractionwatch.com
A flagship study claiming AI chatbots can debunk conspiracy theories has been flagged by Science — exposing just how fragile our trust in "inoculation by algorithm" really is.
Inoculation TheoryBelief PerseveranceEpistemic TrustMisinformation Correction
Theory Briefing
- The September 2024 Science paper claimed DebunkBot reduced belief in specific conspiracy theories — but the journal has now flagged the study for concern.
- The finding fed a wave of optimism about AI as a scalable inoculation tool against misinformation, making the retraction watch particularly consequential.
- If the data don't hold up, it challenges the core premise that brief AI-mediated dialogue can reliably shift deeply held conspiratorial beliefs.