Understanding the efficacy of provenance interventions for tackling misinformation

Abstract

The proliferation of misinformation has fuelled widespread research and societal concern. Several interventions aim to reduce users’ susceptibility. However, audiovisual misinformation such as cheapfakes can be difficult for individuals and fact-checkers to detect and com- bat at scale. Leveraging cryptographic algorithms to track the origins, manipulation, and circulation of audiovisual content, provenance cues offer a novel intervention. This paper conducts the first pilot online experiment (n = 108) to test the efficacy of provenance cues against recontextualized visual misinformation about climate change. Findings offer partial support that provenance cues reduce user-level misinformation susceptibility and have implications that foreground the need for a new research frontier.