Human Detection of Political Deepfakes across Transcripts, Audio, and Video

Abstract

Recent advances in technology for hyper-realistic visual effects provoke the concern that deepfake videos of political speeches will soon be visually indistinguishable from authentic video recordings. Yet there exists little empirical research on how audio-visual information influences people's susceptibility to fall for political misinformation. The conventional wisdom in the field of communication research predicts that people will fall for fake news more often when the same version of a story is presented as a video as opposed to text. However, audio-visual manipulations often leave distortions that some but not all people may pick up on. Here, we evaluate how communication modalities influence people's ability to discern real political speeches from fabrications based on a randomized experiment with 5,727 participants who provide 61,792 truth discernment judgments. We show participants soundbites from political speeches that are randomly assigned to appear using permutations of text, audio, and video modalities. We find that communication modalities mediate discernment accuracy: participants are more accurate on video with audio than silent video, and more accurate on silent video than text transcripts. Likewise, we find participants rely more on how something is said -the audio-visual cues -rather than what is said -the speech content itself. However, political speeches that do not match public perceptions of politicians' beliefs reduce participants' reliance on visual cues. In particular, we find that reflective reasoning moderates the degree to which participants consider visual information: low performance on the Cognitive Reflection Test is associated with an underreliance on visual cues and an overreliance on what is said.

Deepfake Detection in Super-Recognizers and Police Officers

“…Thus, future work should address the extent to which judgements are influenced by prior information on, or familiarity with deepfakes could affect observers' performance -and potentially interact with individual differences in face identity processing ability. Finally, facial deepfakes may be combined with audio content, which in isolation can facilitate or hamper deepfake detection performance [27]. Potentially, the detection of deepfakes involving both audio and visual information could relate to stable individual differences in multisensory integration.…”

What label should be applied to content produced by generative AI?

“…In the context of AIgenerated warning labels, recent work has show that explaining how the label was created can increase the labels' efficacy [5]. For their experiments on the human detection of political deepfakes, [15] include a disclosure stating the content "contains AI-generated content" but do not test the efficacy of that particular disclosure statement.…”

Creating, Using, Misusing, and Detecting Deep Fakes

“…A recent study, however, challenges this conventional wisdom (Groh, Sankaranarayanan, and Picard 2022). After seeing or reading a short political statement by either Joe Biden or Donald Trump, participants were asked to determine if the statement could be attributed to either individual.…”