Average accuracy levels for each video, dis-aggregated for authentic (left) and fake (right) videos. The error bars denote 95% confidence intervals. The red line indicates the 50% accuracy rate of random guessing.
Author: Köbis, NC; Doležalová, B; Soraperra, I
Description: Hyper-realistic manipulations of audio-visual content, i.e., deepfakes, present new challenges for establishing the veracity of online content. Research on the human impact of deepfakes remains sparse. In a pre-registered behavioral experiment (NÂ = 210), we show that (1) people cannot reliably detect deepfakes and (2) neither raising awareness nor introducing financial incentives improves their detection accuracy. Zeroing in on the underlying cognitive processes, we find that (3) people are biased toward mistaking deepfakes as authentic videos (rather than vice versa) and (4) they overestimate their own detection abilities. Together, these results suggest that people adopt a “seeing-is-believing” heuristic for deepfake detection while being overconfident in their (low) detection abilities. The combination renders people particularly susceptible to be influenced by deepfake content.
Subject headings: Neuroscience; Behavioral neuroscience; Cognitive neuroscience; Artificial intelligence; Artificial intelligence applications; Social sciences; Psychology; Deepfakes; Video
Publication year: 2021
Journal or book title: iScience
Volume: 24
Issue: 11
Pages: 103364
Find the full text: https://www.strategian.com/fulltext/Nobis2021.pdf
Find more like this one (cited by): https://scholar.google.com/scholar?cites=18342066305317629739&as_sdt=1000005&sciodt=0,16&hl=en
Serial number: 4177