Providing detection strategies to improve human detection of deepfakes: An experimental study
Open accessAbstract
Deepfake videos are becoming more pervasive. In this preregistered online experiment, participants (N = 454, Mage = 37.19, SDage = 13.25, males = 57.5%) categorize a series of 20 videos as either real or deepfake. All participants saw 10 real and 10 deepfake videos. Participants were randomly assigned to receive a list of strategies for detecting deepfakes based on visual cues (e.g., looking for common artifacts such as skin smoothness) or to act as a control group. Participants were also asked how confident they were that they categorized each video correctly (per video confidence) and to estimate how many videos they correctly categorized out of 20 (overall confidence). The sample performed above chance on the detection activity, correctly categorizing 60.70% of videos on average (SD = 13.00). The detection strategies intervention did not impact detection accuracy or detection confidence, with the intervention and control groups performing similarly on the detection activity and showing similar levels of confidence. Inconsistent with previous research, the study did not find that participants had a bias toward categorizing videos as real. Participants overestimated their ability to detect deepfakes at the individual video level. However, they tended to underestimate their abilities on the overall confidence question.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Somoray & Miller (2023) found mean categorization accuracy of 60.7% for 20 videos lasting 10 seconds, only slightly exceeding chance.
“Similarly, another recent study (Somoray & Miller, 2023) found the mean categorization accuracy of 20 videos lasting 10 seconds to be 60.7%, which, again, only slightly exceeded chance levels.”
From Deepfakes, Democracy, and the Ethics of Synthetic Media: A Synthesis of the SOLARIS Project (2026) · cites this paper as Somoray & Miller, 2023
People generally believe that they can reliably detect deepfakes and overestimate their performance in deepfake detection tasks.
“People generally believe that they can reliably detect deepfakes and overestimate their performance in deepfake detection tasks (e.g., Köbis et al., 2021;Somoray & Miller, 2023)”
From Deepfakes, Democracy, and the Ethics of Synthetic Media: A Synthesis of the SOLARIS Project (2026) · cites this paper as Somoray & Miller, 2023
Human-centered methods often struggle to accurately identify AI-generated content.
“Human-centered methods, which are dependent on individual judgments, often struggle to accurately identify AI-generated content (Groh et al. 2024; Somoray and Miller 2023; Diel et al. 2024; Bray, Johnson, and Kleinberg 2023).”
From Characterizing AI-Generated Misinformation on Social Media (2025) · cites this paper as Somoray and Miller 2023
Providing basic detection guidance (written tips) yielded no improvement on detection accuracy.
“Providing this kind of basic detection guidance yielded no improvement on detection accuracy [4, 6].”
From Human Performance in Deepfake Detection: A Systematic Review (2025) · cites this paper as [4]
Explicitly teaching detection skills by providing written detection tips does not appear to improve detection compared to control.
“Furthermore, explicitly teaching detection skills by providing written detection tips does not appear to improve detection compared to control [4, 6].”
From Human Performance in Deepfake Detection: A Systematic Review (2025) · cites this paper as [4]
The citation lists haven’t been fetched yet
References 20 · Cited by 47
- Authors
- Somoray, K., & Miller, D. J.
- Affiliations
- James Cook University (AU)
- Corresponding author
- Somoray, Klaire
- Venue
- Computers in Human Behavior
- Publisher
- Elsevier BV
- Published
- 2023-12
- Volume
- 149
- Pages
- 107917
- ISSN
- 0747-5632 (ISSN-L)
- OpenAlex
- W4386034004
- Cited by
- 47
- References
- 20
- Citations by year
- 2026: 15 · 2025: 17 · 2024: 14 · 2023: 1
- FWCI
- 4.27
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- Computers in Human Behavior (publishedVersion, cc-by-nc)
- ResearchOnline - JCU (James Cook University) (submittedVersion, other-oa)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Digital Media Forensic Detection
- Social Sciences › Sociology and Political Science › Misinformation and Its Impacts
- Social Sciences › Psychology › Social Psychology › Deception detection and forensic psychology
- Keywords
- deepfakes, deepfake detection, video authenticity, detection methods, visual artifacts, overconfidence
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
