The Deepfake Detection Dilemma: A Multistakeholder Exploration of Adversarial Dynamics in Synthetic Media
Abstract
Synthetic media detection technologies label media as either synthetic or non-synthetic and are increasingly used by journalists, web platforms, and the general public to identify misinformation and other forms of problematic content. As both well-resourced organizations and the non-technical general public generate more sophisticated synthetic media, the capacity for purveyors of problematic content to adapt induces a detection dilemma : as detection practices become more accessible, they become more easily circumvented. This paper describes how a multistakeholder cohort from academia, technology platforms, media entities, and civil society organizations active in synthetic media detection and its socio-technical implications evaluates the detection dilemma. Specifically, we offer an assessment of detection contexts and adversary capacities sourced from the broader, global AI and media integrity community concerned with mitigating the spread of harmful synthetic media. A collection of personas illustrates the intersection between unsophisticated and highly-resourced sponsors of misinformation in the context of their technical capacities. This work concludes that there is no "best'' approach to navigating the detector dilemma, but derives a set of implications from multistakeholder input to better inform detection process decisions and policies, in practice.
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?
Recent advances in technology for hyper-realistic manipulations to video pose a threat to traditional standards of evidence and trust in media.
“Recent advances in technology for algorithmically applying hyper-realistic manipulations to video are simultaneously enabling new forms of interpersonal communication and posing a threat to traditional standards of evidence and trust in media1–8.”
From Human detection of political speech deepfakes across transcripts, audio, and video (2024) · cites this paper as [4]
Cites
Image-to-Image Translation with Conditional Adversarial Networks
Isola, P., Zhu, J.-Y., Zhou, T., & Efros, A. A. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.632 · Cited by 22,930 · see citing works
Media Forensics and DeepFakes: An Overview
Verdoliva, L. (2020) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2020.3002101 · Cited by 979 · see citing works
Artificial Fingerprinting for Generative Models: Rooting Deepfake Attribution in Training Data
Yu, N., Skripniuk, V., Abdelnabi, S., & Fritz, M. (2021) · 2021 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.01418 · Cited by 195 · see citing works
Exposing GAN-synthesized Faces Using Landmark Locations
Yang, X., Li, Y., Qi, H., & Lyu, S. (2019) · ACM Workshop on Information Hiding and Multimedia Security (IH&MMSec) · doi:10.1145/3335203.3335724 · Cited by 127 · see citing works
The Foundational Cryptography Framework
Petcher, A., & Morrisett, G. (2015) · Lecture notes in computer science · doi:10.1007/978-3-662-46666-7_4 · Cited by 81 · see citing works
Encounters with Visual Misinformation and Labels Across Platforms
Saltz, E., Leibowicz, C., & Wardle, C. (2021) · Extended Abstracts of the CHI Conference on Human Factors in Computing Systems (CHI EA) · doi:10.1145/3411763.3451807 · Cited by 61 · see citing works
Self-citationDeepfakes: A Grounded Threat Assessment
Center for Security and Emerging Technology, & Hwang, T. (2020) · doi:10.51593/20190030 · Cited by 33 · see citing works
Flagged first, then by citations
2 of the 9 references couldn’t be matched to a known paper.
Cited by
Sorted by citations
Deepfake detection by human crowds, machines, and machine-informed crowds
Groh, M., Epstein, Z., Firestone, C., & Picard, R. (2022) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2110013119 · Cited by 248 · see citing works
Typology of Risks of Generative Text-to-Image Models
Bird, C., Ungless, E., & Kasirzadeh, A. (2023) · Proceedings of the 2023 AAAI/ACM Conference on AI, Ethics, and Society · doi:10.1145/3600211.3604722 · Cited by 121 · see citing works
Human detection of political speech deepfakes across transcripts, audio, and video
Groh, M., Sankaranarayanan, A., Singh, N., Kim, D. Y., Lippman, A., & Picard, R. (2024) · Nature Communications · doi:10.1038/s41467-024-51998-z · Cited by 89 · see citing works
Weikmann, T., & Lecheler, S. K. (2023) · Digital Journalism · doi:10.1080/21670811.2023.2194665 · Cited by 49 · see citing works
Browne, T. O., Abedin, M., & Chowdhury, M. J. M. (2024) · International Journal of Information Security · doi:10.1007/s10207-024-00868-2 · Cited by 34 · see citing works
Txt2Vid: Ultra-Low Bitrate Compression of Talking-Head Videos via Text
Tandon, P., Chandak, S., Pataranutaporn, P., Liu, Y., Mapuranga, A. M., Maes, P., Weissman, T., & Sra, M. (2022) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2022.3221953 · Cited by 32 · see citing works
The spread of synthetic media on X
Corsi, G., Marino, B., & Wong, W. (2024) · Harvard Kennedy School Misinformation Review · doi:10.37016/mr-2020-140 · Cited by 24 · see citing works
Navigating legal challenges of deepfakes in the American context: a call to action
Chawki, M. (2024) · Cogent Engineering · doi:10.1080/23311916.2024.2320971 · Cited by 22 · see citing works
The impact of advanced AI systems on democracy
Summerfield, C., Argyle, L. P., Bakker, M. A., Collins, T., Durmus, E., Eloundou, T., Gabriel, I., Ganguli, D., Hackenburg, K., Hadfield, G. K., et al. (2025) · Nature Human Behaviour · doi:10.1038/s41562-025-02309-z · Cited by 41 · see citing works
Self-citationDomenteanu, A., Tătaru, G.-C., Crăciun, L., Molănescu, A. G., Cotfas, L., & Delcea, C. (2024) · Information · doi:10.3390/info15090525 · Cited by 17 · see citing works
DeepExplain: Enhancing DeepFake Detection Through Transparent and Explainable AI model
Sunkari, V., & Srinagesh, A. (2024) · Informatica · doi:10.31449/inf.v48i8.5792 · Cited by 18 · see citing works
Deepfake AI images: should deepfakes be banned in Thailand?
Songja, R., Promboot, I., Haetanurak, B., & Kerdvibulvech, C. (2023) · AI and Ethics · doi:10.1007/s43681-023-00350-0 · Cited by 11 · see citing works
SpecViT: A Custom Vision-Transformer based Approach for Audio Deepfake Detection
Modak, S., Das, A. K., & Naskar, R. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10889022 · Cited by 7 · see citing works
Goldstein, J. A., & DiResta, R. (2022) · Harvard Kennedy School Misinformation Review · doi:10.37016/mr-2020-104 · Cited by 7 · see citing works
Ethical Aspects of Content Creation
Patalauskaitė, E. (2024) · Filosofija Sociologija · doi:10.6001/fil-soc.2024.35.3.14 · Cited by 5 · see citing works
Yıldız, G. (2025) · İmgelem · doi:10.53791/imgelem.1732361 · Cited by 2 · see citing works
Rao, T. V. N., Koppula, M., Vardhan, G. H., Bhavana, S., & Kumar, B. S. (2024) · Advances in business information systems and analytics book series · doi:10.4018/979-8-3693-6890-9.ch011 · Cited by 1 · see citing works
Synthetic Artwork Authentication Threats: Detection by Combining Neural Network and Blockchain
Kearns, L., Alam, A., & Allison, J. (2025) · Transactions on Emerging Telecommunications Technologies · doi:10.1002/ett.70225 · Cited by 1 · see citing works
Information Consumption Patterns, Fake News, and Deep Fake
Kumar, S., Verma, A. K., & Mirza, A. (2024) · Frontiers of artificial intelligence, ethics and multidisciplinary applications. · doi:10.1007/978-981-97-5656-8_9
Garg, H., Kashif, M., Sharma, P., & Malik, A. (2025) · Information systems engineering and management · doi:10.1007/978-3-031-90742-5_12
An Intelligent AI-Generated Content Detection Tool and Framework
Annapurna, S., Manjunath, G. V. S., Ram, K. A. S. S., Ashwini, B., & Vardhan, D. H. (2026) · International Conference on Machine Learning and Autonomous Systems (ICMLAS) · doi:10.1109/icmlas67792.2026.11483921
Yeredla, D., Archana, S., Rangu, V., Shirisha, R., & Varun, V. S. (2026) · IEEE Global Symposium on Emerging and Communication Technologies (GSEACT) · doi:10.1109/gseact68539.2026.11620577
- Authors
- Leibowicz, C. R., McGregor, S., & Ovadya, A.
- Affiliations
- The Partnership on AI, San Francisco, CA, USA
- XPRIZE Foundation (US)
- Thoughtful Technology Project, San Francisco, CA, USA
- Venue
- Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society
- Publisher
- ACM
- Published
- 2021-07-21
- Pages
- 736-744
- ISSN
- 3065-8365 (ISSN-L)
- OpenAlex
- W3186090001
- Cited by
- 23
- References
- 9
- Citations by year
- 2026: 2 · 2025: 5 · 2024: 9 · 2023: 3 · 2022: 2 · 2021: 2
- FWCI
- 10.67
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- Proceedings of the AAAI/ACM Conference on AI Ethics and Society (publishedVersion)
- Topics
- Social Sciences › Sociology and Political Science › Misinformation and Its Impacts
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Digital Media Forensic Detection
- Social Sciences › Safety Research › Ethics and Social Impacts of AI
- Keywords
- deepfake detection, synthetic media, misinformation, media integrity, multi-stakeholder perspectives, personas
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
