Deepfake Warnings for Political Videos Increase Disbelief but Do Not Improve Discernment: Evidence from Two Experiments
Open accessAbstract
Recent advances in machine learning have led to the development of the “deepfake,” a convincingly realistic, computer-generated video of a public figure saying something they have not actually said. Policymakers have expressed concern that deepfakes could mislead voters and affect election outcomes, but existing research has found minimal persuasive effects. In this paper, we explore a downstream consequence of deepfakes: if voters are repeatedly warned of the existence and dangers of deepfakes, they may simply begin to distrust all political video footage – whether real or fake. Through two online survey experiments, we found that voters were unable to discriminate between a real video and a deepfake. Statements warning about the existence of deepfakes did not enhance participants’ ability to successfully spot manipulated video content. Instead, these warnings consistently induced participants to believe that the videos they watched were fake, even when the videos were real. The warnings were not specific to the video participants were watching; simply stating that deepfakes exist increased distrust of any accompanying video. Our findings suggest that even if deepfakes are not themselves persuasive, rhetoric about deepfakes can nevertheless be weaponized by politicians and campaigns to dismiss and disown real videos.
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?
People's agreement with the content of a message significantly affects judgments of the authenticity of videos.
“Studies have shown the importance of assessing message content itself as people ' s agreement with the content of a message signi fi cantly a ff ects judgments of the authenticity of videos [38, 39].”
From Human Performance in Deepfake Detection: A Systematic Review (2025) · cites this paper as [39]
Providing warning labels may cause participants to doubt the authenticity of all videos watched, regardless of whether they are fake or not.
“However, providing warning labels may have the unintended consequence of causing participants to doubt the authenticity of all videos watched, regardless of whether they fake or not [39].”
From Human Performance in Deepfake Detection: A Systematic Review (2025) · cites this paper as [39]
Providing warnings about the potential for a video being deepfaked increases mistrust of authentic videos.
“while providing warnings about the potential for a video being deepfaked seems to increase detection of deepfakes, but at the expense of mistrust of authentic videos [39].”
From Human Performance in Deepfake Detection: A Systematic Review (2025) · cites this paper as [39]
Providing warnings about the potential for a video being deepfaked seems to increase detection of deepfakes.
“while providing warnings about the potential for a video being deepfaked seems to increase detection of deepfakes, but at the expense of mistrust of authentic videos [39].”
From Human Performance in Deepfake Detection: A Systematic Review (2025) · cites this paper as [39]
Cites
The spread of true and false news online
Vosoughi, S., Roy, D., & Aral, S. (2018) · Science · doi:10.1126/science.aap9559 · Cited by 8,847 · see citing works
Lazer, D. M. J., Baum, M. A., Benkler, Y., Berinsky, A. J., Greenhill, K. M., Menczer, F., Metzger, M. J., Nyhan, B., Pennycook, G., Rothschild, D., et al. (2018) · Science · doi:10.1126/science.aao2998 · Cited by 4,160 · see citing works
Pennycook, G., & Rand, D. G. (2019) · Cognition · doi:10.1016/j.cognition.2018.06.011 · Cited by 2,022 · see citing works
Pennycook, G., McPhetres, J., Zhang, Y., Lu, J. G., & Rand, D. G. (2020) · Psychological Science · doi:10.1177/0956797620939054 · Cited by 1,715 · see citing works
Field Experiments: Design, Analysis, and Interpretation
GREEN, D. P. (2024) · doi:10.61700/sgeirzx0v0mjh1252 · Cited by 1,322 · see citing works
Coppock, A., & McClellan, O. A. (2019) · Research & Politics · doi:10.1177/2053168018822174 · Cited by 987 · see citing works
Vaccari, C., & Chadwick, A. (2020) · Social Media + Society · doi:10.1177/2056305120903408 · Cited by 978 · see citing works
Pennycook, G., Bear, A., Collins, E., & Rand, D. G. (2020) · Management Science · doi:10.1287/mnsc.2019.3478 · Cited by 718 · see citing works
Partial Identification of Probability Distributions
(2003) · Springer Series in Statistics · doi:10.1007/b97478 · Cited by 698 · see citing works
Clayton, K., Blair, S., Busam, J. A., Forstner, S., Glance, J., Green, G., Kawata, A., Kovvuri, A., Martin, J., Morgan, E., et al. (2020) · Political Behavior · doi:10.1007/s11109-019-09533-0 · Cited by 658 · see citing works
Follow the Leader?: How Voters Respond to Politicians' Policies and Performance
Lenz, G. (2012) · Cited by 603 · see citing works
Priming and Fake News: The Effects of Elite Discourse on Evaluations of News Media
Duyn, E. V., & Collier, J. R. (2018) · Mass Communication & Society · doi:10.1080/15205436.2018.1511807 · Cited by 427 · see citing works
Pennycook, G., McPhetres, J., Zhang, Y., Lu, J. G., & Rand, D. G. (2020) · doi:10.31234/osf.io/uhbk9 · Cited by 390 · see citing works
Do (microtargeted) deepfakes have real effects on political attitudes?
Dobber, T., Metoui, N., Trilling, D., Helberger, N., & de Vreese, C. (2021) · The International Journal of Press/Politics · doi:10.1177/1940161220944364 · Cited by 256 · see citing works
“Don’t Know” Means “Don’t Know”: DK Responses and the Public’s Level of Political Knowledge
Luskin, R. C., & Bullock, J. G. (2011) · The Journal of Politics · doi:10.1017/s0022381611000132 · Cited by 249 · see citing works
The Question(s) of Political Knowledge
Barabas, J., Jerit, J., Pollock, W. M., & Rainey, C. (2014) · American Political Science Review · doi:10.1017/s0003055414000392 · Cited by 240 · see citing works
The Generalizability of Online Experiments Conducted During the COVID-19 Pandemic
Peyton, K., Huber, G. A., & Coppock, A. (2021) · Journal of Experimental Political Science · doi:10.1017/xps.2021.17 · Cited by 174 · see citing works
Field experiments design, analysis, and interpretation
Gerber, A. S., & Green, D. P. (2012) · Cited by 149 · see citing works
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 · see citing works
Evidence of Rising Rates of Inattentiveness on Lucid in 2020Preprint
Aronow, P. M., Kalla, J., Orr, L., & Ternovski, J. (2020) · doi:10.31235/osf.io/8sbe4 · Cited by 106 · see citing works
Self-citationAvoiding the Hypothetical: Why “Mirror Experiments” are an Essential Part of Survey Research
McDonald, J. A. (2019) · International Journal of Public Opinion Research · doi:10.1093/ijpor/edz027 · Cited by 76 · see citing works
The Generalizability of Online Experiments Conducted During The COVID-19 PandemicPreprint
Peyton, K., Huber, G. A., & Coppock, A. (2020) · doi:10.31235/osf.io/s45yg · Cited by 69 · see citing works
Understanding and Reducing the Spread of Misinformation Online
Pennycook, G., Epstein, Z. G., Mosleh, M., Arechar, A. A., Eckles, D., & Rand, D. G. (2020) · ACR North American Advances · Cited by 61 · see citing works
Flagged first, then by citations
1 of the 24 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
Martel, C., & Rand, D. G. (2023) · Current Opinion in Psychology · doi:10.1016/j.copsyc.2023.101710 · Cited by 122 · see citing works
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 · see citing works
“As a result, it is perhaps unsur-prising that multimodal misinformation is often viewed as highly credible (38–40)—especially relative to textual misinfor-mation (refs.”
Negative Downstream Effects of Alarmist Disinformation Discourse: Evidence from the United States
Jungherr, A. M., & Rauchfleisch, A. (2024) · Political Behavior · doi:10.1007/s11109-024-09911-3 · Cited by 53 · see citing works
“…exposure to misleading information (Altay et al., 2022 ; Vaccari & Chad - wick, 2020 ), exposure to general warnings about misleading information (Ternovski et al., 2021 ; Van Duyn & Collier, 2019 ), exposure to labels indicating specific infor - mational items to be false (Jahng et al., 2021),…”
Sadiq, S., Aljrees, T., & Ullah, S. (2023) · IEEE Access · doi:10.1109/access.2023.3308515 · Cited by 56 · see citing works
“created problems in a number of fields such as politics [4] by deceiving viewers into thinking that they were created by humans.”
BackgroundReview of social science research on the impact of countermeasures against influence operations
Courchesne, L., Ilhardt, J. G., & Shapiro, J. N. (2021) · Harvard Kennedy School Misinformation Review · doi:10.37016/mr-2020-79 · Cited by 48 · see citing works
“While some papers have found that partisan, ideologically congruent misinformation is particularly resistant to change (i.e., Nyhan & Reifler, 2010; Walter & Salovich, 2021), others did not identify a partisan effect (Nyhan & Reifler, 2016).”
BackgroundUsing deepfakes for experiments in the social sciences - A pilot study
Eberl, A., Kühn, J., & Wolbring, T. (2022) · Frontiers in Sociology · doi:10.3389/fsoc.2022.907199 · Cited by 48 · see citing works
“Experiments using manipulated videos point in the same direction corroborating the claim that people cannot reliably detect deepfakes (Khodabakhsh et al., 2019; Köbis et al., 2021; Ternovski et al., 2021).”
ResultDigital Technology, Politics, and Policy-Making
Gilardi, F. (2022) · Cambridge University Press eBooks · doi:10.1017/9781108887304 · Cited by 41 · see citing works
Martel, C., & Rand, D. G. (2023) · doi:10.31234/osf.io/48p2a · Cited by 30 · see citing works
Political Deepfakes Are As Credible As Other Fake Media And (Sometimes) Real Media
Barari, S., Lucas, C., & Munger, K. (2021) · doi:10.31219/osf.io/cdfh3 · Cited by 51 · see citing works
Regulating Deep Fakes in the Artificial Intelligence Act
Łabuz, M. (2023) · Applied Cybersecurity & Internet Governance · doi:10.60097/acig/162856 · Cited by 29 · see citing works
Kreps, S., & Kriner, D. L. (2024) · New Media & Society · doi:10.1177/14614448231160526 · Cited by 27 · see citing works
“Deepfakes have proliferated online in ways that can impugn the reputation of public officials and celebrities and are increasingly able to evade human detection (Ternovski et al., 2021).”
OECD (2020) · OECD skills outlook · doi:10.1787/e11c1c2d-en · Cited by 26 · see citing works
Human Performance in Deepfake Detection: A Systematic Review
Somoray, K., Miller, D. J., & Holmes, M. (2025) · Human Behavior and Emerging Technologies · doi:10.1155/hbe2/1833228 · Cited by 19 · see citing works
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 15 · see citing works
“Some studies have found that participants have no greater overall accuracy in judging fake news in the presence of tags (Ternovski et al., 2021).”
BackgroundAI or Your Lying Eyes: Some Shortcomings of Artificially Intelligent Deepfake Detectors
Harris, K. R. (2024) · Philosophy & Technology · doi:10.1007/s13347-024-00700-8 · Cited by 13 · see citing works
“Adding to this concern is recent empirical evidence suggesting that warnings about deep-fakes themselves reduce willingness to believe based on even non-deepfake video footage—in other words, that such warnings constitute a skeptical threat in their own right (Ternovski et al., 2021).”
Ruffin, M., Seo, H., Xiong, A., & Wang, G. (2024) · Proceedings of the International AAAI Conference on Web and Social Media · doi:10.1609/icwsm.v18i1.31392 · Cited by 8 · see citing works
“The results showed participants cannot effectively differentiate them (Ternovski, Kalla, and Aronow 2021; Groh et al. 2022; Korshunov and Marcel 2020).”
BackgroundResultKang, S., & Valadez, K. (2025) · Journalism & Mass Communication Quarterly · doi:10.1177/10776990251357294 · Cited by 8 · see citing works
“As Ternovski et al. (2021) point out, however, indiscriminate deepfake warnings can erode trust in the information the audience views, causing a deception default.”
Prediction of self-efficacy in recognizing deepfakes based on personality traits
Abraham, J., Putra, H. A., Prayoga, T., Warnars, H. L. H. S., Manurung, R. H., & Nainggolan, T. (2022) · F1000Research · doi:10.12688/f1000research.128915.1 · Cited by 7 · see citing works
Impact of Deepfake Technology on FinTech Applications
Abbas, N. N., Ahmad, R., Qazi, S., & Ahmed, W. (2022) · Advances in information security, privacy, and ethics book series · doi:10.4018/978-1-6684-5284-4.ch012 · Cited by 5 · see citing works
Deepfake Characterization, Propagation, and Detection in Social Media - A Synthesis Review
Sudarsan, A., Chua, H. N., Jasser, M. B., & Wong, R. T. K. (2024) · IEEE International Colloquium on Signal Processing & Its Applications (CSPA) · doi:10.1109/cspa60979.2024.10525373 · Cited by 5 · see citing works
“However, a contradictory factor, with [25] showing a decline in detection accuracy with age, found younger users more likely to share fake news.”
BackgroundA Pilot Study Investigating STEM Learners’ Ability to Decipher AI-generated Video
Shu, D., Doss, C., Mondschein, J., Kopecky, D., Fitton-Kane, V. A., Bush, L. B., & Tucker, C. S. (2021) · ASEE Annual Conference & Exposition · doi:10.18260/1-2--36601 · Cited by 5 · see citing works
“These findings align with previous studies that have found large amounts of uncertainty among individuals asked to discern between deepfake and authentic digital media, although these literature results are mixed regarding which content – deepfake or authentic – results in greater uncertainty [31], [47].”
ResultPrediction of self-efficacy in recognizing deepfakes based on personality traits
Abraham, J., Putra, H. A., Prayoga, T., Warnars, H. L. H. S., Manurung, R. H., & Nainggolan, T. (2023) · F1000Research · doi:10.12688/f1000research.128915.3 · Cited by 2 · see citing works
“A person ’ s Honesty-humility trait do not want to engineer others but, ironically, this trait makes them vulnerable to being manipulated by others (Ternovski et al. , 2021), including deepfakes.”
Khalid, M., Mushtaq, M. F., Akram, U., Safran, M., Alfarhood, S., & Ashraf, I. (2025) · Scientific Reports · doi:10.1038/s41598-025-10661-3 · Cited by 2 · see citing works
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 2 · see citing works
AST: Generalization of Deepfake Detection with Attention Siamese Training
Peng, T., Wang, T., Liu, D., Wang, J., Fu, Y., & Snoussi, H. (2023) · China Automation Congress (CAC) · doi:10.1109/cac59555.2023.10451796 · Cited by 1 · see citing works
“At present, the internet is flooded with malicious videos produced by deepfake programs, posing grave threats to personal privacy and societal stability [1,2].”
BackgroundOnebunne, A. P. (2025) · International Journal of Arts and Humanities · doi:10.25082/ijah.2025.01.003 · Cited by 1 · see citing works
The structural fingerprints of disinformation: a content-agnostic framework
Redaelli, S., Spitale, G., & Germani, F. (2026) · Ethics and Information Technology · doi:10.1007/s10676-025-09886-7 · Cited by 1 · see citing works
“Psychosocial interventions are strategies designed to influence an individual’s mindset or engage their in-group and/or social identity (Blair et al., 2024).”
O IMPACTO DISRUPTIVO QUE O DEEPFAKE PODE CAUSAR NA SOCIEDADE
Silva, E. F., Casalenovo, L. D. B., & Amaral, C. A. (2023) · Revista Rumos da Pesquisa em Ciências Empresarias Ciências do Estado e da Tecnologia · doi:10.17648/2525-278x-v1n7-6 · Cited by 1 · see citing works
Deepfake on Social Media: Harnessing Deep Learning for Identifying Falsified Tweet
Kaur, J. K. (2024) · INTERANTIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT · doi:10.55041/ijsrem32156 · Cited by 1 · see citing works
The Use of Deep Fakes in Social Engineering Attacks
Saxena, A., & Grewal, H. (2024) · Advances in information security, privacy, and ethics book series · doi:10.4018/979-8-3693-6665-3.ch013 · Cited by 1 · see citing works
Sadiq, S., & Ullah, S. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4494619 · Cited by 1 · see citing works
Prediction of self-efficacy in recognizing deepfakes based on personality traits
Abraham, J., Putra, H. A., Prayoga, T., Warnars, H. L. H. S., Manurung, R. H., & Nainggolan, T. (2023) · F1000Research · doi:10.12688/f1000research.128915.2
Deepfake, Disinformation and Social Media
Bhatta, S. (2024) · International Journal for Research in Applied Science and Engineering Technology · doi:10.22214/ijraset.2024.61625
Assessing the Believability of Deepfake Ads in the Indian Market
Menon, V. A., & Gupta, A. (2024) · Advances in business information systems and analytics book series · doi:10.4018/979-8-3693-6890-9.ch003
Risbakk, J., & Skogstad, M. R. (2026) · Communications in computer and information science · doi:10.1007/978-3-032-17272-3_17
Vadapalli, S. K., Doyal, S., Vanheusden, F. J., Binder, J. F., & Kuss, D. J. (2026) · SN Social Sciences · doi:10.1007/s43545-026-01349-0
Balcı, Ş., & Özben, K. (2026) · Kritik İletişim Çalışmaları Dergisi · doi:10.53281/kritik.1877790
- Authors
- Ternovski, J., Kalla, J., & Aronow, P. M.
- Affiliations
- Yale University (US)
- Publisher
- Center for Open Science
- Published
- 2021-01-14
- OpenAlex
- W4232480655
- Cited by
- 37
- References
- 24
- Citations by year
- 2026: 4 · 2025: 5 · 2024: 8 · 2023: 11 · 2022: 4 · 2021: 4 · 2020: 1
- Type
- posted content
- Language
- en
- Open access
- green
- doi.org
- Licence
- Locations
- OSF Preprints (OSF Preprints) (submittedVersion, public-domain)
- OSF Preprints (OSF Preprints) (submittedVersion)
- Funders
- Yale University
- Topics
- Social Sciences › Sociology and Political Science › Misinformation and Its Impacts
- Social Sciences › Sociology and Political Science › Media Influence and Politics
- Social Sciences › Political Science and International Relations › Electoral Systems and Political Participation
- Keywords
- deepfakes, political videos, video manipulation detection, political campaigns, disinformation, voter persuasion
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
