puc-header

Fooled Twice – People Cannot Detect Deepfakes But Think They Can

30 Pages Posted: 23 Apr 2021 Publication Status: Published

See all articles by Nils Köbis

Nils Köbis

Max Planck Institute for Human Development - Center for Humans and Machines

Barbora Doležalová

Max Planck Society for the Advancement of the Sciences - Center for Humans and Machines

Ivan Soraperra

University of Amsterdam - Center for Research in Experimental Economics and Political Decision-Making (CREED)

More...

Abstract

Hyper-realistic manipulation of audio-visual content, i.e., deepfakes, presents a new challenge for establishing veracity of online content. Research on the human impact of deepfakes, addressing both behaviors in response to and cognitive processing of deepfakes, remains sparse. In a pre-registered behavioral experiment ( N = 210), we show that (a) people cannot reliably detect deepfakes, and (b) neither raising awareness nor introducing financial incentives improves their detection accuracy. Zeroing in on the underlying cognitive processes, we find that (c) people are biased towards mistaking deepfakes as authentic videos (rather than vice versa) and (d) 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 inauthentic deepfake content.

Keywords: Deepfake detection, Experiment, Machine Behavior, Overcon dence

Suggested Citation

Köbis, Nils and Doležalová, Barbora and Soraperra, Ivan, Fooled Twice – People Cannot Detect Deepfakes But Think They Can. Available at SSRN: https://ssrn.com/abstract=3832978 or http://dx.doi.org/10.2139/ssrn.3832978
This version of the paper has not been formally peer reviewed.

Nils Köbis (Contact Author)

Max Planck Institute for Human Development - Center for Humans and Machines ( email )

Berlin
Germany

HOME PAGE: http://https://www.mpib-berlin.mpg.de/person/107772

Barbora Doležalová

Max Planck Society for the Advancement of the Sciences - Center for Humans and Machines

Berlin
Germany

Ivan Soraperra

University of Amsterdam - Center for Research in Experimental Economics and Political Decision-Making (CREED) ( email )

Faculty of Economics and Econometrics
Roetersstraat 11
Amsterdam, 1018 WB
Netherlands

Click here to go to Cell.com

Paper statistics

Abstract Views
1,537
Downloads
51
PlumX Metrics