CoChat

Testing human ability to detect ‘deepfake’ images of human faces

Open access
Bray, Sergi D, Johnson, Shane D, Kleinberg, Bennett · 2023 · Journal of Cybersecurity

The DOI matches its Crossref record; 98 of the 98 references it cites were found in a registry or an index.

Abstract

‘Deepfakes’ are computationally created entities that falsely represent reality. They can take image, video, and audio modalities, and pose a threat to many areas of systems and societies, comprising a topic of interest to various aspects of cybersecurity and cybersafety. In 2020, a workshop consulting AI experts from academia, policing, government, the private sector, and state security agencies ranked deepfakes as the most serious AI threat. These experts noted that since fake material can propagate through many uncontrolled routes, changes in citizen behaviour may be the only effective defence. This study aims to assess human ability to identify image deepfakes of human faces (these being uncurated output from the StyleGAN2 algorithm as trained on the FFHQ dataset) from a pool of non-deepfake images (these being random selection of images from the FFHQ dataset), and to assess the effectiveness of some simple interventions intended to improve detection accuracy. Using an online survey, participants (N = 280) were randomly allocated to one of four groups: a control group, and three assistance interventions. Each participant was shown a sequence of 20 images randomly selected from a pool of 50 deepfake images of human faces and 50 images of real human faces. Participants were asked whether each image was AI-generated or not, to report their confidence, and to describe the reasoning behind each response. Overall detection accuracy was only just above chance and none of the interventions significantly improved this. Of equal concern was the fact that participants’ confidence in their answers was high and unrelated to accuracy. Assessing the results on a per-image basis reveals that participants consistently found certain images easy to label correctly and certain images difficult, but reported similarly high confidence regardless of the image. Thus, although participant accuracy was 62% overall, this accuracy across images ranged quite evenly between 85 and 30%, with an accuracy of below 50% for one in every five images. We interpret the findings as suggesting that there is a need for an urgent call to action to address this threat.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1093/cybsec/tyad011

Cites

98 of 98 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 2 Example of highlighting explanatory regions of image during experiment.
Figure 2 Example of highlighting explanatory regions of image during experiment.CC-BY · From the PDF
Figure 3 Use of 'tell-tale signs' in participants' reasoning behind response decisions, by decision correctness (where participants' responses contained 'Deepfake' as the selected label).
Figure 3 Use of 'tell-tale signs' in participants' reasoning behind response decisions, by decision correctness (where participants' responses contained 'Deepfake' as the selected label).CC-BY · From the PDF
Image 3
CC-BY · From the PDF
Image 4
CC-BY · From the PDF

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

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?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 98 references

Research works printed but not linked: 33

  1. Merriam Webster . Words w e're watc hing: "Deepfak e." 2020. https:// web.archive.org/ web/ 20230314205626/https:// www.merria m-webster.com/words-at-play/deepfake-slang-definition-examples , (7 June 2023, date last accessed).
  2. Conor . Prescriptivist vs descriptivist-what's the difference? One Minute English . 2021. https:// web.archive.org/ web/ 20230320155207/https:// oneminuteenglish.org/en/ prescriptivist-descriptivist/ (7 June 2023, date last accessed).
  3. Foster Wallace D. Authority and american usage. In: Consider The Lob- ster . New York, NY: Little, Brown and Company, 2006, 66-127.
  4. Deepf ake , W ikipedia. https:// web.archive.org/ web/ 20230603085859/h ttps:// en.wikipedia.org/wiki/ Deepfake (7 June 2023, date last accessed).
  5. Dash B, Sharma P. Are ChatGPT and deepfake algorithms endangering the cybersecurity industry? A review. Int J Appl Sci Eng 2023; 10 :1-5.
  6. Style Transfer Guide | Fritz AI. https:// www.fritz.ai/style-transfer/ (7 June 2023, date last accessed).
  7. DALL •E 2. OpenAI . https:// web.archive.org/ web/ 20230606194002/htt ps://openai.com/dall-e-2/ (7 June 2023, date last accessed).
  8. Damiani J. A Voice Deepfake Was Used To Scam A CEO Out Of $243,000. Forbes 2019. https:// web.archive.org/ web/ 20230511004754 /https:// www.forbes.com/sites/jessedamiani/ 2019/09/ 03/a-voice-deepf ake-was-used-to-scam-a-ceo-out-of-243000/ (7 June 2023, date last accessed).
  9. Snapshot Paper-Deepfakes and Audiovisual Disinformation. GOVUK 2020. https://www .gov .uk/government/publications/cdei-publishes-its- first-series-of-three-snapshot-papers-ethical-issues-in-ai/snapshot-pap er-deepfakes-and-audiovisual-disinformation (7 June 2023, date last accessed).
  10. The Hypnotic Allure of the AI Art Generator. Sothebys.com 2019. https:// www.sothebys.com/en/ videos/the-hypnotic-allure-of-the-ai-art -generator (7 June 2023, date last accessed).
  11. VICE News . How Does A.I. Art Stack Up Against Human Art? (HBO) , 2018. https:// www.vice.com/en/ article/ ywqw8j/how-does-ai-a rt-stack-up-against-human-art (7 June 2023, date last accessed).
  12. Galvani M. You Shall Be. SoundCloud 2021. https://soundcloud.com/m g273/you-shall-be/s-Tpqs2595DlQ (7 June 2023, date last accessed).
  13. Oest A, Zhang P, Wardman B. et al. Sunrise to Sunset: Analyzing the End-to-end Life Cycle and Effectiveness of Phishing Attacks at Scale. 29th {USENIX} Security Symposium ({USENIX} Security 20) , 2020: 18 .
  14. The Draft Online Safety Bill and the legal but harmful de- bate Date Published: 24 January 2022 UK Parliament. GOVUK 2020. https:// publications.parliament.uk/pa/ cm5802/cmselect/ cmcume ds/1039/report.html (7 June 2023, date last accessed).
  15. Gow G. The Scary Truth Behind the FBI Warning: Deepfake Fraud is Here And it's Serious-We Are Not Prepared. Forbes 2021. https:// web.archive.org/ web/ 20221207202324/https:// www.forbes.com /sites/glenngow/ 2021/05/ 02/the-scary-truth-behind-the-fbi-warning- deepfake-fraud-is-here-and-its-serious-we-are-not-prepared/ (7 June 2023, date last accessed).
  16. The Grandparent Scam. Federal Bureau of Inv estig ation , 2021. https:// web.archive.org/ web/ 20230321235226/https:// www.fbi.gov/ ne ws/stories/the-grandparent-scam (7 June 2023, date last accessed).
  17. Albahar M, Almalki J. Deepfakes: threats and Countermeasures Sys- tematic Review. J Theor Appl Inf Technol 2019; 97 :3242-50.
  18. Reuters Institute, 2020, Digital News Report 2020
  19. Reuters Institute, 2018, Digital News Report 2018
  20. The Strategy and Tactics of the Pro-Kremlin Disinformation Campaign. EU vs Disinfo 2018. https:// web.archive.org/ web/ 20230523071717/htt ps://euvsdisinfo.eu/the-strategy-and-tactics-of-the-pro-kremlin-disinfo rmation-campaign/ (7 June 2023, date last accessed).
  21. Toxic Twitter -Women's Experiences of Violence and Abuse on Twitter. 2018. https:// web.archive.org/ web/ 20230429193706/https: //www .amnesty .org/en/latest/news/2018/03/online-violence-against-w omen-chapter-3-2/ (7 June 2023, date last accessed).
  22. Great Britain, Committee on Standards in Public Life . Intimidation in Public Life: A Review by the Committee on Standards in Public Life . 2017. https://www .gov .uk/government/publications/intimidation-in-p ublic-life-a-review-by-the-committee-on-standards-in-public-life (7 June 2023, date last accessed).
  23. Stevens R. How blockchain could help combat deepfakes. Decrypt 2019. https:// web.archive.org/ web/ 20230604120533/https:// decrypt.co /11588/how-blockchain-could-help-combat-deepfakes/ (7 June 2023, date last accessed).
  24. Rashid MM, Lee S-H, Kwon K-R. Blockchain technology for combat- ing deepfake and protect video/image integrity. J. Korea Multimed. Soc. 2021; 24 :1044-58. https:// www.ncbi.nlm.nih.gov/pmc/articles/ PMC918 5340/#B15-sensors-22-04256 .
  25. Sato G. What Is Account Takeover Fraud and How Can You Prevent It?. Experian 2020. https:// web.archive.org/ web/ 20230201003001/https: // www.experian.com/blogs/ask-experian/ what-is-account-takeover-fra ud-how-to-prevent-it/ (7 June 2023, date last accessed).
  26. Guerra PHC, Guedes D, Meira JW. et al. Exploring the Spam Arms Race to Characterize Spam Evolution. Proceedings of the 7th Collabo- ration, Electronic messaging, Anti-Abuse and Spam Conference (CEAS), Redmond, WA , 2010.
  27. 2020, NVlabs/Ffhq-Dataset
  28. Metz, 2019, These people do not exist. Why websites are churning out fake images of people (and cats), CNN
  29. O'Sullivan, 2019, Now fake Facebook accounts are using fake faces, CNN
  30. O'Sullivan, 2020, How fake faces are being weaponized online, CNN
  31. Vincent, 2019, A spy reportedly used an AI-generated profile picture to connect with sources on LinkedIn, The Verge
  32. 2021, Deepest-Project/MelNet
  33. 2021, This MP Does Not Exist, VOLE.wtf

Cited by

Metadata

Authors
Bray, S. D., Johnson, S. D., & Kleinberg, B.
Affiliations
  • University College London (GB)
  • Tilburg University (NL)
Corresponding author
Bray, Sergi D
Venue
Journal of Cybersecurity
Published
2023-01-01
Volume
9
Issue
1
ISSN
  • 2057-2085 (ISSN-L)
  • 2057-2093
OpenAlex
W4381851764
Cited by
96
References
95
Citations by year
2026: 27 · 2025: 34 · 2024: 30 · 2023: 5
FWCI
8.73
Type
journal article
Language
en
Open access
Locations
  • Journal of Cybersecurity (publishedVersion, cc-by)
  • Research portal (Tilburg University) (submittedVersion, cc-by)
Funders
EPSRC (EP/S022503/1)
Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Digital Media Forensic Detection
  • Social Sciences › Sociology and Political Science › Misinformation and Its Impacts
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Face Recognition and Perception
Keywords
deepfakes, deepfake detection, face images, StyleGAN2, confidence calibration, misinformation, AI risks, cybersecurity

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

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.