Political Deepfakes Are As Credible As Other Fake Media And (Sometimes) Real Media
Open accessThe DOI matches its Crossref record; 94 of the 94 references it cites were found in a registry or an index.
Abstract
We demonstrate that fabricated videos of public officials synthesized by deep learning ("deepfakes'") are credible to a large portion of the American public - up to 50% of a representative sample of 5,750 subjects - however no more than equivalent misinformation in extant modalities like text headlines or audio recordings. Moreover, there are no meaningful heterogeneities in these credibility perceptions nor greater affective responses relative to other mediums across subgroups. However, when asked to discern real videos from deepfakes, partisanship explains a large gap in viewers' detection accuracy, but only for real videos, not deepfakes. Brief informational messages or accuracy primes only sometimes (and somewhat) attenuate deepfakes' effects. Above all else, broader literacy in politics and digital technology increases discernment between deepfakes and authentic videos of political elites. Our findings come from two experiments testing exposure to a novel collection of deepfakes created in collaboration with tech industry partners.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
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?
In the citing paper
“When faced with deepfake content of politicians, participants rely on contextual knowledge in the form of political literacy to identify spoofs [20, 21].”
From Warning: Humans cannot reliably detect speech deepfakes (2023) · cites this paper as [20]
Citations
Cites
Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing
Benjamini, Y., & Hochberg, Y. (1995) · Journal of the Royal Statistical Society Series B: Statistical Methodology · doi:10.1111/j.2517-6161.1995.tb02031.x · Cited by 109,482 · cited 2×
“Thicker lines denote 95% confidence intervals (CIs), thinner lines denote 95% CIs adjusted for multiple subgroup comparisons (Benjamini and Hochberg, 1995), red blocks indicate 95% CIs from two one-sided equivalence t-tests with equivalence bounds indicated by the vertical red lines (Wellek, 2010).”
BackgroundMethodInfluentialAn Economic Theory of Democracy.
Banfield, E. C., & Downs, A. (1958) · Midwest Journal of Political Science · doi:10.2307/2109186 · Cited by 18,594
Understanding Media: The Extensions of Man
Levine, S., & McLuhan, M. (1964) · American Quarterly · doi:10.2307/2711172 · Cited by 12,311
TensorFlow: A system for large-scale machine learning
Abadi, M., Barham, P., Chen, J., Chen, Z., Davis, A., Dean, J., Devin, M., Ghemawat, S., Irving, G., Isard, M., et al. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1605.08695 · Cited by 11,822
Manufacturing Consent: The Political Economy of the Mass Media.
Bennett, J. R., Herman, E. S., & Chomsky, N. (1989) · Contemporary Sociology · doi:10.2307/2074220 · Cited by 5,437
The Ambivalent Sexism Inventory: Differentiating hostile and benevolent sexism.
Glick, P., & Fiske, S. T. (1996) · Journal of Personality and Social Psychology · doi:10.1037//0022-3514.70.3.491 · Cited by 5,202 · cited 1×
“3Ambivalent sexism describes a bundle of both outright hostile (e.g. “women are physically inferior to men”) and deceptively benevolent views about women (e.g. “women are objects of desire”) (Glick and Fiske, 1996) 4 susceptible to deepfake misinformation over text and audio misinformation.”
BackgroundAnomalous collapses of Nares Strait ice arches leads to enhanced export of Arctic sea ice
Moore, G. W. K., Howell, S. E. L., Brady, M., Xu, X., & McNeil, K. (2021) · Nature Communications · doi:10.1038/s41467-020-20314-w · Cited by 5,155
What Americans Know about Politics and Why It Matters
Brooks, C., Carpini, M. X. D., & Keeter, S. (1997) · Contemporary Sociology · doi:10.2307/2655098 · Cited by 4,437
Pennycook, G., & Rand, D. G. (2019) · Cognition · doi:10.1016/j.cognition.2018.06.011 · Cited by 2,022 · cited 3×
“N on -In te rv en ab le in Su rv ey Sexists (with female target) • Consistency with prior hostile beliefs about women (Glick and Fiske, 1996; Schaffner, MacWilliams and Nteta, 2018; Cassese and Holman, 2019) • Consistency with prior benevolent beliefs about women (Glick and Fiske, 1996; Schaffner, MacWilliams and Ntet…”
BackgroundAmerican journal of political science
(1983) · Electoral Studies · doi:10.1016/0261-3794(83)90076-8 · Cited by 1,985
Cited by
The psychological drivers of misinformation belief and its resistance to correction
Ecker, U. K. H., Lewandowsky, S., Cook, J., Schmid, P., Fazio, L. K., Brashier, N., Kendeou, P., Vraga, E. K., & Amazeen, M. A. (2022) · Nature Reviews Psychology · doi:10.1038/s44159-021-00006-y · Cited by 1,372
Pawelec, M. (2022) · Digital Society · doi:10.1007/s44206-022-00010-6 · Cited by 131
The (minimal) persuasive advantage of political video over text
Wittenberg, C., Tappin, B. M., Berinsky, A. J., & Rand, D. G. (2021) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2114388118 · Cited by 114
Warning: Humans cannot reliably detect speech deepfakes
Mai, K. T., Bray, S., Davies, T., & Griffin, L. D. (2023) · PLOS ONE · doi:10.1371/journal.pone.0285333 · Cited by 98
“[20] mention fake speech stimuli in their analysis.”
BackgroundInfluentialQian, S., Shen, C., & Zhang, J. (2022) · Journal of Computer-Mediated Communication · doi:10.1093/jcmc/zmac024 · Cited by 99
Human performance in detecting deepfakes: A systematic review and meta-analysis of 56 papers
Diel, A., Lalgi, T., Schröter, I. C., MacDorman, K. F., Teufel, M., & Bäuerle, A. (2024) · Computers in Human Behavior Reports · doi:10.1016/j.chbr.2024.100538 · Cited by 64
They Would Never Say Anything Like This! Reasons To Doubt Political Deepfakes
Hameleers, M., van der Meer, T. G. L. A., & Dobber, T. (2024) · European Journal of Communication · doi:10.1177/02673231231184703 · Cited by 47
Labeling AI-generated media online
Wittenberg, C., Epstein, Z., Péloquin-Skulski, G., Berinsky, A. J., & Rand, D. G. (2025) · PNAS Nexus · doi:10.1093/pnasnexus/pgaf170 · Cited by 45
“Following (27), we adopted an active debriefing strategy, wherein respondents had to explicitly declare (in writing) that they understood the content they were shown was fake.”
The Liar’s Dividend: Can Politicians Claim Misinformation to Evade Accountability?
Schiff, K. J., Schiff, D. S., & Bueno, N. S. (2024) · American Political Science Review · doi:10.1017/s0003055423001454 · Cited by 98
“On the other hand, other recent studies call into question the extent to which deepfakes are more credible and persuasive than text-based misinformation (Barari, Lucas, and Munger 2021), though video-based misinformation may be more effective for changing beliefs (Wittenberg et al. 2021).”
InfluentialKaate, I., Salminen, J. O., Santos, J. M., Jung, S., Olkkonen, R., & Jansen, B. J. J. (2023) · International Journal of Human-Computer Studies · doi:10.1016/j.ijhcs.2023.103096 · Cited by 46
How News Coverage of Misinformation Shapes Perceptions and Trust
Thorson, E. A. (2024) · Cambridge University Press eBooks · doi:10.1017/9781009488815 · Cited by 42
Fighting Infodemics: Labels as Antidotes to Mis- and Disinformation?!
Mende, M., Ubal, V. O., Cozac, M., Vallen, B., & Berry, C. (2023) · Journal of Public Policy & Marketing · doi:10.1177/07439156231184816 · Cited by 39
“For instance, people ’ s disposition to cognitive re fl ection affects the detection of falsi fi ed content (Barari, Lucas, and Munger 2021).”
BackgroundThe Use of AI by Election Campaigns
Foos, F. (2024) · LSE Public Policy Review · doi:10.31389/lseppr.112 · Cited by 23
Kaate, I., Salminen, J. O., Jung, S., Almerekhi, H., & Jansen, B. J. J. (2023) · Biannual Conference of the Italian SIGCHI Chapter (CHItaly) · doi:10.1145/3605390.3605397 · Cited by 44
Walker, C., Schiff, D. S., & Schiff, K. J. (2024) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v38i21.30349 · Cited by 46
Kaate, I., Salminen, J. O., Santos, J. M., Jung, S., Almerekhi, H., & Jansen, B. J. J. (2023) · Computers in Human Behavior Artificial Humans · doi:10.1016/j.chbah.2023.100031 · Cited by 31
KARAKOÇ, E., & ZEYBEK, B. (2021) · Öneri Dergisi · doi:10.14783/maruoneri.908542 · Cited by 23
Artificial Intelligence in Election Campaigns: Perceptions, Penalties, and Implications
Jungherr, A. M., Rauchfleisch, A., & Wuttke, A. (2026) · Political Communication · doi:10.1080/10584609.2025.2611913 · Cited by 41
“…of political figures using deepfakes, and the deployment of AI-generated content in coordinated astroturfing campaigns (e.g., via bots or automated e-mail outreach to journalists and voters) (e.g., Barari et al., 2025; Bond, 2024; Coltin, 2024; Linvill & Warren, 2024; Ternovski et al., 2022).”
Deepfakes: Vehicles for Radicalization, Not Persuasion
Nieweglowska, M., Stellato, C., & Sloman, S. A. (2023) · Current Directions in Psychological Science · doi:10.1177/09637214231161321 · Cited by 16
“Even more convincing, Köbis et al. (2021) show that people cannot reliably detect deepfakes and that neither raising awareness nor introducing financial incentives improves their detection accuracy.”
BackgroundExploding AI-Generated Deepfakes and Misinformation: A Threat to Global Concern in the 21st CenturyPreprint
Singh, P. (2023) · doi:10.36227/techrxiv.24715605.v1 · Cited by 20
Media Forensics in the Age of Disinformation
Hendrix, J., & Morozoff, D. (2022) · Advances in computer vision and pattern recognition · doi:10.1007/978-981-16-7621-5_2 · Cited by 16
MacKenzie, W. I., Weber, R. P., Barr, H. M., Lanius, C. L., & Tenhundfeld, N. L. (2025) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2025.2466068 · Cited by 13
The Liar’s Dividend: Can Politicians Claim Misinformation to Evade Accountability?
Schiff, K. J., Schiff, D. S., & Bueno, N. (2023) · doi:10.31235/osf.io/x43ph · Cited by 8
Kang, S., & Valadez, K. (2025) · Journalism & Mass Communication Quarterly · doi:10.1177/10776990251357294 · Cited by 12
“One positive sign is that viewers are likely to notice disinformation in the video, given their cognitive ability and prior knowledge (Barari et al., 2024).”
von Sikorski, C., & Hameleers, M. (2025) · Journalism & Mass Communication Quarterly · doi:10.1177/10776990251375097 · Cited by 27
“To date, different experimental studies suggest that AI-powered disinformation, mostly studied in the context of deepfake videos, does not yield stronger effects than other forms of disinformation, such as textual disinformation (e.g., Barari et al., 2025).”
Exploding AI-Generated Deepfakes and Misinformation: A Threat to Global Concern in the 21st CenturyPreprint
Singh, P., & Dhiman, B. (2023) · doi:10.36227/techrxiv.24715605 · Cited by 16
The harm of deepfakes: a scoping review of deepfakes’ negative effects on human mind and behavior
Diel, A., Lalgi, T. J., Mellis, F. S., Teufel, A., & Bäuerle, A. (2025) · AI & Society · doi:10.1007/s00146-025-02774-0 · Cited by 23
“Empirical evidence shows that political deepfakes can appear credible (Appel and Prietzel 2022; Barari et al. 2021; 2025), especially when the message is consistent with the consumers’ beliefs (Hameleers et al. 2022).”
Dhahir, D. F., Kenda, N., Dirgahayu, D., Syarifuddin, S., Djaffar, R., Widihastuti, R., Rustam, M., & Pala, R. (2024) · Jurnal Pekommas · doi:10.56873/jpkm.v9i2.5873 · Cited by 22
The Liar’s Dividend: Can Politicians Use Deepfakes and Fake News to Evade Accountability?
Schiff, K. J., Schiff, D. S., & Bueno, N. (2022) · doi:10.31235/osf.io/q6mwn · Cited by 6
The Problem of Misinformation and Disinformation Online
Rubin, V. L. (2022) · Springer eBooks · doi:10.1007/978-3-030-95656-1_1 · Cited by 5
Exploding AI-Generated Deepfakes and Misinformation: A Threat to Global Concern in the 21st CenturyPreprint
Dhiman, B. (2023) · Qeios · doi:10.32388/dple2l · Cited by 10
“By synchronizing different modalities, these deepfakes become harder to detect and more persuasive [10] .”
BackgroundThe overlooked threat of democratic neutrality in the USA
Hall, M. E. K., Leigh, B. T., & Solomon, B. C. (2026) · Nature Human Behaviour · doi:10.1038/s41562-026-02430-7 · Cited by 3
Hameleers, M., & van der Meer, T. G. L. A. (2026) · New Media & Society · doi:10.1177/14614448251409208 · Cited by 5
Oliullah, M., & Murtuza, H. M. (2026) · Computers in Human Behavior Reports · doi:10.1016/j.chbr.2026.101066 · Cited by 6
Wack, M., & Prochaska, S. (2026) · Social Media + Society · doi:10.1177/20563051261462092 · Cited by 5
“Consistent with prior work (Barari et al., 2021; Hameleers et al., 2022; Vaccari & Chadwick, 2020), our results suggest that deceptive deepfakes can harm trust in targeted candidates.”
A framework for studying causal effects of speech style: application to US presidential campaigns
Damann, T. J., Knox, D., & Lucas, C. (2025) · Journal of the Royal Statistical Society Series A (Statistics in Society) · doi:10.1093/jrsssa/qnaf059 · Cited by 3
Self-citationSun, Y.-B., Shen, L.-W., Choi, J.-S., Borah, P., Wagner, M. W., Shah, D. V., & Yang, S.-J. (2025) · Journalism & Mass Communication Quarterly · doi:10.1177/10776990251359659 · Cited by 2
“Journalism & Mass Communication Quarterly 102(4) and fabricated content (Barari et al., 2025; Q. Peng et al., 2024).”
Exploring the Impact of Deepfakes on Cognitive Processing in Political Contexts: A Systematic ReviewPreprint
Oliullah, M., & Murtuza, H. M. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5390735 · Cited by 4
Hameleers, M., van der Meer, T. G. L. A., Tulin, M., & Dobber, T. (2026) · Communication Research · doi:10.1177/00936502261421437 · Cited by 7
Still Instrumentally Inclusive
Turnbull-Dugarte, S. J., & Ortega, A. L. (2026) · American Political Science Review · doi:10.1017/s0003055426101828 · Cited by 1
“See, for example, Baisley and Albaugh (2025), Barari, Lucas, and Munger (2025), Broockman and Kalla (2025), Culpepper, Jung, and Lee (2024), Grose, Malhotra, and Van Houweling (2015), Hainmueller and Hopkins (2015), Marsh and Peterson (2026), Matanock and García-Sánchez (2018), Mikkelborg (2025),…”
Deepfakes and Synthetic Media: Generation, Detection, and Governance
Gazis, A., Karypidis, E., Santamouri, K., Vavouras, T., Mastorakis, N. E., & Pappas, S. S. (2026) · Encyclopedia · doi:10.3390/encyclopedia6080165 · Cited by 5
“This is important if we consider that confirmation bias has a marked impact on user perception [15].”
InfluentialWe Are Equally Vulnerable? An Examination of the Third-Person Effect in AI-Generated Misinformation
Fu, B., & Li, X. (2026) · Social Media + Society · doi:10.1177/20563051261452019
Strategies for Recognising Deepfake Videos in the Development of Media Literacy
Vrabec, N., & Hoti, V. (2025) · Media & Marketing Identity · doi:10.34135/mmidentity-2025-70 · Cited by 1
Bi, C., & Zhu, Y. (2026) · Cogent Arts and Humanities · doi:10.1080/23311983.2026.2696098
Villacura-Herrera, C., Zaal, F. T. J. M., & Cox, R. F. A. (2026) · Computers in Human Behavior Reports · doi:10.1016/j.chbr.2026.101221
Linegar, M., Sinclair, B., van der Linden, S., & Alvarez, R. M. (2026) · Royal Society Open Science · doi:10.1098/rsos.252226 · Cited by 1
KÖSELİÖREN, M. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-2600-1242-0.ch004
Does Listening to Trustworthy News Affect Political Attitudes and Trust?
Barberá, P., Guess, A. M., Munzert, S., & Yang, J. (2026) · Political Behavior · doi:10.1007/s11109-026-10193-0
The newspaper Pravda as a phenomenon of the digital transformation of a party publication
Plotnikov, I. P., & Parkhitko, N. P. (2026) · Litera · doi:10.25136/2409-8698.2026.9.81436
Securing the vote: a longitudinal assessment of US election cybersecurity, 2012–2024
Park, C. H., Oh, J., & Reeds, M. (2026) · Journal of Cyber Policy · doi:10.1080/23738871.2026.2737128
Mendieta, D., & Lopez, L. A. G. (2026) · REVISTA DE DERECHO · doi:10.21703/issn2735-6337/2026.n49.04
Metadata
- Authors
- Barari, S., Lucas, C., & Munger, K.
- Affiliations
- University of Chicago (US)
- Washington University in St. Louis (US)
- European University Institute (IT)
- Venue
- The Journal of Politics
- Publisher
- Center for Open Science
- Published
- 2021-01-13
- Volume
- 87
- Issue
- 2
- Pages
- 510-526
- ISSN
- 0022-3816 (ISSN-L)
- OpenAlex
- W4402554561
- Cited by
- 51
- References
- 60
- Citations by year
- 2026: 17 · 2025: 9 · 2024: 6 · 2023: 11 · 2022: 6 · 2021: 2
- FWCI
- 22.87
- Type
- journal article
- Language
- en
- Open access
- green
- hdl.handle.net
- Licence
- other-oa
- Locations
- The Journal of Politics (publishedVersion)
- Cadmus - EUI Research Repository (European University Institute) (submittedVersion, other-oa)
- OSF Preprints (OSF Preprints) (submittedVersion)
- Topics
- Social Sciences › Sociology and Political Science › Misinformation and Its Impacts
- Social Sciences › Sociology and Political Science › Media Influence and Politics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Digital Media Forensic Detection
- Keywords
- political deepfakes, misinformation, media credibility, deception detection, video authenticity, partisan polarization, media trust, information interventions
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.