Research can help to tackle AI-generated disinformation
Open accessThe claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
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?
Automated detection approaches often struggle with accuracy and robustness, leading to false positives or undetected deepfakes.
“However, they often struggle with accuracy and robustness, leading to false positives or undetected deepfakes (Almars 2021; Feuerriegel et al. 2023; Zhou et al. 2023).”
From Characterizing AI-Generated Misinformation on Social Media (2025) · cites this paper as Feuerriegel et al. 2023
The “Regulatory and technological mitigations” section says AI-content detection tools have notable shortcomings, including false positives, and predicts that evolving generators will probably outpace detectors. It does not specifically report undetected deepfakes as a current outcome, so that part of the claim is more definite and specific than the source. (LLM, conf=0.82)
Experimental — accuracy not yet measuredThe manipulation of media content has long been a concern; however, rapid advancements of generative AI tools have significantly lowered the barriers to creating sophisticated AI-generated content.
“The manipulation of media content has long been a concern; however, the rapid advancements of generative AI tools have significantly lowered the barriers to creating sophisticated AI-generated content (Westerlund 2019; Feuerriegel et al. 2023).”
From Characterizing AI-Generated Misinformation on Social Media (2025) · cites this paper as Feuerriegel et al. 2023
The increased scalability, multilingualism, and multimodality of AI-generated content further complicate its detection and mitigation.
“Additionally, the increased scalability, multilingualism, and multimodality of AI-generated content further complicate its detection and mitigation (Feuerriegel et al. 2023; Timmerman et al. 2023).”
From Characterizing AI-Generated Misinformation on Social Media (2025) · cites this paper as Feuerriegel et al. 2023
AI-generated misinformation poses significant challenges to defense strategies previously employed by digital platforms and users.
“This emerging form of digital misinformation poses significant challenges to the defense strategies previously employed by digital platforms and users (Goldstein et al. 2023; Feuerriegel et al. 2023).”
From Characterizing AI-Generated Misinformation on Social Media (2025) · cites this paper as Feuerriegel et al. 2023
LLMs are adept at impersonating humans at scale, making malicious use to spread misinformation in the short term very concerning.
“The potential for malicious actors to use Generative AI to spread misinformation in the short term is very concerning because tools such as LLMs are so adept at impersonating humans at scale (67).”
From Can Generative AI improve social science? (2024) · cites this paper as [67]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜
Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021) · Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3442188.3445922 · Cited by 6,988 · see citing works
Citizens Versus the Internet: Confronting Digital Challenges With Cognitive Tools
Kozyreva, A., Lewandowsky, S., & Hertwig, R. (2020) · Psychological Science in the Public Interest · doi:10.1177/1529100620946707 · Cited by 479 · see citing works
All the News That’s Fit to Fabricate: AI-Generated Text as a Tool of Media Misinformation
Kreps, S., McCain, R. M., & Brundage, M. (2022) · Journal of Experimental Political Science · doi:10.1017/xps.2020.37 · Cited by 360 · see citing works
AI model GPT-3 (dis)informs us better than humans
Spitale, G., Biller-Andorno, N., & Germani, F. (2023) · Science Advances · doi:10.1126/sciadv.adh1850 · Cited by 241 · see citing works
How behavioural sciences can promote truth, autonomy and democratic discourse online
Lorenz-Spreen, P., Lewandowsky, S., Sunstein, C. R., & Hertwig, R. (2020) · Nature Human Behaviour · doi:10.1038/s41562-020-0889-7 · Cited by 223 · see citing works
Self-citationGoldstein, J. A., Sastry, G., Musser, M., DiResta, R., Gentzel, M., & Sedova, K. (2023) · arXiv · doi:10.48550/arxiv.2301.04246 · Cited by 152 · see citing works
Self-citationHe, B., Ahamad, M., & Kumar, S. (2023) · International World Wide Web Conference (WWW) · doi:10.1145/3543507.3583388 · Cited by 49 · see citing works
Self-citationPolitical audience diversity and news reliability in algorithmic ranking
Bhadani, S., Yamaya, S., Flammini, A., Menczer, F., Ciampaglia, G. L., & Nyhan, B. J. (2022) · Nature Human Behaviour · doi:10.1038/s41562-021-01276-5 · Cited by 42 · see citing works
New deepfake regulations in China are a tool for social stability, but at what cost?
Hine, E., & Floridi, L. (2022) · Nature Machine Intelligence · doi:10.1038/s42256-022-00513-4 · Cited by 36 · see citing works
Kreps, S., & Kriner, D. L. (2024) · New Media & Society · doi:10.1177/14614448231160526 · Cited by 27 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
The impact of generative artificial intelligence on socioeconomic inequalities and policy making
Capraro, V., Lentsch, A., Acemoglu, D., Akgun, S., Akhmedova, A., Bilancini, E., Bonnefon, J.-F., Brañas-Garza, P., Butera, L., Douglas, K. M., et al. (2024) · PNAS Nexus · doi:10.1093/pnasnexus/pgae191 · Cited by 371 · see citing works
Can Generative AI improve social science?
Bail, C. A. (2024) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2314021121 · Cited by 369 · see citing works
We 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
“Currently, research on AI-generated misinformation primarily focuses on its characteristics, such as automated production, high realism (Feuerriegel et al., 2023), and its prevalence in entertainment-related topics (Drolsbach & Pröllochs, 2025).”
Characterizing AI-Generated Misinformation on Social Media
Drolsbach, C., Demirel, E., & Pröllochs, N. (2025) · arXiv · doi:10.48550/arxiv.2505.10266
Self-citation
- Authors
- Feuerriegel, S., DiResta, R., Goldstein, J. A., Kumar, S., Lorenz-Spreen, P., Tomz, M., & Pröllochs, N.
- Affiliations
- Munich Center for Machine Learning (DE)
- Stanford University (US)
- Georgetown University (US)
- Georgia Institute of Technology (US)
- Max Planck Institute for Human Development (DE)
- Stanford Institute for Economic Policy Research (US)
- Justus-Liebig-Universität Gießen (DE)
- Corresponding author
- Feuerriegel, Stefan
- ORCID
- Venue
- Nature Human Behaviour
- Published
- 2023-11-20
- Volume
- 7
- Issue
- 11
- Pages
- 1818-1821
- ISSN
- 2397-3374 (ISSN-L)
- OpenAlex
- W4388834782
- Cited by
- 64
- References
- 10
- Citations by year
- 2026: 21 · 2025: 23 · 2024: 18 · 2023: 2
- FWCI
- 27.58
- Type
- journal article
- Language
- en
- Open access
- Licence
- Locations
- Nature Human Behaviour (publishedVersion)
- PubMed (publishedVersion)
- Open access LMU (Ludwid Maxmilian's Universitat Munchen) (submittedVersion)
- RePEc: Research Papers in Economics (submittedVersion)
- MPG.PuRe (Max Planck Society) (submittedVersion, other-oa)
- Topics
- Social Sciences › Sociology and Political Science › Misinformation and Its Impacts
- Physical Sciences › Computer Science › Information Systems › Spam and Phishing Detection
- Social Sciences › Safety Research › Ethics and Social Impacts of AI
- Keywords
- AI-generated disinformation, generative AI, disinformation detection, online information reliability, public trust, research agenda, misinformation
- MeSH terms
- Disinformation, Artificial Intelligence, Humans
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
