Evaluating Twitter’s algorithmic amplification of low-credibility content: an observational study
Open accessAbstract
Artificial intelligence (AI)-powered recommender systems play a crucial role in determining the content that users are exposed to on social media platforms. However, the behavioural patterns of these systems are often opaque, complicating the evaluation of their impact on the dissemination and consumption of disinformation and misinformation. To begin addressing this evidence gap, this study presents a measurement approach that uses observed digital traces to infer the status of algorithmic amplification of low-credibility content on Twitter over a 14-day period in January 2023. Using an original dataset of ≈ 2.7 million posts on COVID-19 and climate change published on the platform, this study identifies tweets sharing information from low-credibility domains, and uses a bootstrapping model with two stratifications, a tweet’s engagement level and a user’s followers level, to compare any differences in impressions generated between low-credibility and high-credibility samples. Additional stratification variables of toxicity, political bias, and verified status are also examined. This analysis provides valuable observational evidence on whether the Twitter algorithm favours the visibility of low-credibility content, with results indicating that, on aggregate, tweets containing low-credibility URL domains perform better than tweets that do not across both datasets. However, this effect is largely attributable to a difference in high-engagement, high-followers tweets, which are very impactful in terms of impressions generation, and are more likely receive amplified visibility when containing low-credibility content. Furthermore, high toxicity tweets and those with right-leaning bias see heightened amplification, as do low-credibility tweets from verified accounts. Ultimately, this suggests that Twitter’s recommender system may have facilitated the diffusion of false content by amplifying the visibility of low-credibility content with high-engagement generated by very influential users.
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.
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?
Studies observe increased low-credibility content in algorithmic timelines, with right-leaning users experiencing higher exposure to such content.
“Studies also observe increased low-credibility content in algorithmic timelines [10], with right-leaning users experiencing higher exposure to such content [9].”
From Auditing Political Exposure Bias: Algorithmic Amplification on Twitter/X During the 2024 U.S. Presidential Election (2025) · cites this paper as [10]
Section 3.2 (page unavailable) reports greater amplification of right-leaning low-credibility tweets, which supports increased visibility of that content. The citation is attached to a claim about right-leaning users, however; the paper classifies the political bias of content, not the political leaning of users receiving it, so it does not establish that those users had higher exposure. (LLM, conf=0.86)
Experimental — accuracy not yet measured
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
An Introduction to the Bootstrap
Efron, B., & Tibshirani, R. J. (1994) · doi:10.1201/9780429246593 · Cited by 40,500 · see citing works
The echo chamber effect on social media
Cinelli, M., De Francisci Morales, G., Galeazzi, A., Quattrociocchi, W., & Starnini, M. (2021) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2023301118 · Cited by 2,092 · cited 1× · see citing works
“The emergence of social media platforms has brought about a significant transformation in global patterns of information dissemination and consumption, as a large number of internet users now rely on these channels as their primary sources of information acquisition [1, 2, 3, 4].”
Supervised and Unsupervised Discretization of Continuous Features
Dougherty, J., Kohavi, R., & Sahami, M. (1995) · Elsevier eBooks · doi:10.1016/b978-1-55860-377-6.50032-3 · Cited by 1,946 · see citing works
Emotion shapes the diffusion of moralized content in social networks
Brady, W. J., Wills, J. A., Jost, J. T., Tucker, J. A., & Van Bavel, J. J. (2017) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1618923114 · Cited by 1,396 · cited 1× · see citing works
“These results support findings from prior works, particularly those showing clear connections between negative emotional content and viral misinformation spread [47].”
The filter bubble: how the new personalized web is changing what we read and how we think
(2012) · Choice Reviews Online · doi:10.5860/choice.50-0926 · Cited by 1,319 · see citing works
ChatGPT outperforms crowd workers for text-annotation tasks
Gilardi, F., Alizadeh, M., & Kubli, M. (2023) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2305016120 · Cited by 1,197 · see citing works
Coronavirus Goes Viral: Quantifying the COVID-19 Misinformation Epidemic on Twitter
Kouzy, R., Jaoude, J. A., Kraitem, A. M., Alam, M. B. E., Karam, B. S., Adib, E., Zarka, J. G., Traboulsi, C., Akl, E. A., & Baddour, K. (2020) · Cureus · doi:10.7759/cureus.7255 · Cited by 900 · see citing works
Influence of fake news in Twitter during the 2016 US presidential election
Bovet, A., & Makse, H. A. (2019) · CUNY Academic Works (City University of New York) · Cited by 882 · see citing works
Influence of fake news in Twitter during the 2016 US presidential election
Bovet, A., & Makse, H. A. (2019) · Nature Communications · doi:10.1038/s41467-018-07761-2 · Cited by 701 · see citing works
Auditing radicalization pathways on YouTube
Ribeiro, M. H., Ottoni, R., West, R., Almeida, V. A. F., & Meira, W. (2020) · Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency · doi:10.1145/3351095.3372879 · Cited by 580 · see citing works
Growing polarization around climate change on social media
Falkenberg, M., Galeazzi, A., Torricelli, M., Marco, N. D., Larosa, F., Sas, M. I., Mekacher, A., Pearce, W., Zollo, F., Quattrociocchi, W., et al. (2022) · Nature Climate Change · doi:10.1038/s41558-022-01527-x · Cited by 401 · see citing works
A review of bootstrap confidence intervals
Diciccio, T. J., & Romano, J. P. (1988) · Journal of the Royal Statistical Society Series B: Statistical Methodology · doi:10.1111/j.2517-6161.1988.tb01732.x · Cited by 314 · see citing works
Recommending what video to watch next
Zhao, Z., Hong, L., Li, W., Chen, J., Nath, A., Andrews, S., Kumthekar, A., Sathiamoorthy, M., Yi, X., & Ed, H. (2019) · ACM Conference on Recommender Systems (RecSys) · doi:10.1145/3298689.3346997 · Cited by 314 · see citing works
From “information” to “knowing”: Exploring the role of social media in contemporary news consumption
Pentina, I., & Tarafdar, M. (2014) · Computers in Human Behavior · doi:10.1016/j.chb.2014.02.045 · Cited by 274 · see citing works
Proceedings of the 16th ACM Conference on Recommender Systems
(2022) · doi:10.1145/3523227 · Cited by 219 · see citing works
Link recommendation algorithms and dynamics of polarization in online social networks
Santos, F. P., Yphtach Lelkes, B., & Levin, S. A. (2021) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2102141118 · Cited by 216 · see citing works
academictwitteR: an R package to access the Twitter Academic Research Product Track v2 API endpoint
Barrie, C., & Ho, J. (2021) · Journal of Open Source Software · doi:10.21105/joss.03272 · Cited by 212 · see citing works
A New Generation of Perspective API: Efficient Multilingual Character-level Transformers
Lees, A., Tran, V. Q., Tay, Y., Sorensen, J., Gupta, J., Metzler, D., & Vasserman, L. (2022) · Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining · doi:10.1145/3534678.3539147 · Cited by 155 · see citing works
Social Media Recommender Systems: Review and Open Research Issues
Anandhan, A., Shuib, L. M., Ismail, M. A., & Mujtaba, G. (2018) · IEEE Access · doi:10.1109/access.2018.2810062 · Cited by 142 · see citing works
Comparison of Bootstrap Confidence Interval Methods for GSCA Using a Monte Carlo Simulation
Jung, K., Lee, J., Gupta, V., & Cho, G. (2019) · Frontiers in Psychology · doi:10.3389/fpsyg.2019.02215 · Cited by 114 · see citing works
The Structure of Toxic Conversations on Twitter
Saveski, M., Roy, B., & Roy, D. K. (2021) · International World Wide Web Conference (WWW) · doi:10.1145/3442381.3449861 · Cited by 88 · see citing works
One Year of COVID-19 Vaccine Misinformation on Twitter: Longitudinal Study
Pierri, F., Deverna, M. R., Yang, K.-C., Axelrod, D., Bryden, J., & Menczer, F. (2022) · Journal of Medical Internet Research · doi:10.2196/42227 · Cited by 82 · see citing works
(2021) · doi:10.1145/3442188.3445933 · Cited by 44 · see citing works
Giansiracusa, N. (2021) · doi:10.1007/978-1-4842-7155-1 · Cited by 42 · see citing works
The Soap Box as a Black Box: Regulating transparency in social media recommender systemsPreprint
Leerssen, P. (2020) · doi:10.31228/osf.io/uhxcv · Cited by 27 · see citing works
High level of correspondence across different news domain quality rating setsPreprint
Lin, H., Lasser, J., Lewandowsky, S., Cole, R., Gully, A., Rand, D. G., & Pennycook, G. (2022) · doi:10.31234/osf.io/qy94s · Cited by 20 · see citing works
Flagged first, then by citations
2 of the 28 references couldn’t be matched to a known paper.
Research works printed but not linked: 11
- Islam R, Keya KN, Pan S, Foulds J (2019) Mitigating demographic biases in social media-based recommender systems. KDD (Social Impact Track)
- Alatawi F, Cheng L, Tahir A, Karami M, Jiang B, Black T, Liu H (2021) A survey on echo chambers on social media: Description, detection and mitigation. arXiv preprint arXiv:2112.05084 · arXiv
- Liu Z, Zou L, Zou X, Wang C, Zhang B, Tang D, Zhu B, Zhu Y, Wu P, Wang K (2022) Monolith: real time recommendation system with collisionless embedding table. arXiv preprint arXiv:2209.07663 · arXiv
- Wang Z, She Q, Zhang J (2021) Masknet: introducing feature-wise multiplication to ctr ranking models by instance-guided mask. arXiv preprint arXiv:2102.07619 · arXiv
- Pfeffer J, Matter D, Jaidka K, Varol O, Mashhadi A, Lasser J, Assenmacher D, Wu S, Yang D, Brantner C (2023) Just another day on Twitter: a complete 24 hours of Twitter data. In: Proceedings of the international AAAI conference on web and social media, vol 17, pp 1073-1081
- Treen KMd, Williams HT, O'Neill SJ (2020) Online misinformation about climate change. Wiley Interdiscip Rev: Clim Change 11(5):665
- Brennen JS, Simon FM, Howard PN, Nielsen RK (2020) Types, sources, and claims of covid-19 misinformation. Thesis
- Cuthbertson L, Kearney A, Dawson R, Zawaduk A, Cuthbertson E, Gordon-Tighe A, Mathewson KW (2019) Women, politics and Twitter: using machine learning to change the discourse. arXiv preprint arXiv:1911.11025 · arXiv
- Törnberg P (2023) Chatgpt-4 outperforms experts and crowd workers in annotating political twitter messages with zero-shot learning. arXiv preprint arXiv:2304.06588 · arXiv
- Haroon M, Chhabra A, Liu X, Mohapatra P, Shafiq Z, Wojcieszak M (2022) Youtube, the great radicalizer? auditing and mitigating ideological biases in youtube recommendations. arXiv preprint arXiv:2203.10666 · arXiv
- Ovadya A, Thorburn L (2023) Bridging systems: open problems for countering destructive divisiveness across ranking, recommenders, and governance. arXiv preprint arXiv:2301.09976 · arXiv
Cited by
Sorted by citations
Theodorakopoulos, L., Theodoropoulou, A., & Klavdianos, C. (2025) · Journal of theoretical and applied electronic commerce research · doi:10.3390/jtaer20020115 · Cited by 97 · see citing works
Essien, E. O. (2025) · Social Sciences · doi:10.3390/socsci14050304 · Cited by 33 · see citing works
Ye, J., Luceri, L., & Ferrara, E. (2025) · Proceedings of the 2025 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3715275.3732159 · Cited by 16 · see citing works
“Previous studies consistently show that X ’s algorithm amplifies political biases and prioritizes high-engagement content, including emotionally charged, toxic, and low-credibility information [4, 6, 8–10 , 20].”
Halpern, C., Halpern, B., & Aydın, H. (2025) · Journal of Interdisciplinary Research in Artificial Intelligence and Society · doi:10.20897/jirais/17647 · Cited by 16 · see citing works
Investigating the Influence of AI in Social Media Governance
Khare, P. (2025) · Advances in social networking and online communities book series · doi:10.4018/979-8-3693-9904-0.ch015 · Cited by 11 · see citing works
A crisis of credibility: the global cost of US vaccine misinformation
Larson, H. J., & Piatek, S. J. (2025) · The Lancet · doi:10.1016/s0140-6736(25)01495-3 · Cited by 9 · see citing works
Bacsu, J.-D., Mah, J. C., Jamali, A. A., Conanan, C., Lautrup, S., Berger, C., Fiske, D., Fraser, S., Virani, A., Fehr, F. C., et al. (2025) · JMIR Formative Research · doi:10.2196/72775 · Cited by 7 · see citing works
Naranjo-Vinueza, A., Martín, S. C., González, M. C., & Moreno, P. T. N.-B. (2025) · Frontiers in Political Science · doi:10.3389/fpos.2025.1666104 · Cited by 7 · see citing works
Allen, L. K., DeLiema, D., & Kendeou, P. (2025) · Educational Psychologist · doi:10.1080/00461520.2025.2588213 · Cited by 7 · see citing works
Hausmann, A., Väisänen, T., Toivonen, T., & Cortés-Capano, G. (2025) · Sustainability Science · doi:10.1007/s11625-025-01639-1 · Cited by 7 · see citing works
Ye, J., Luceri, L., & Ferrara, E. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.5018879 · Cited by 6 · see citing works
Wu, C., Jiang, S., Sun, J., & Liu, Y.-Q. (2025) · Acta Psychologica · doi:10.1016/j.actpsy.2025.105560 · Cited by 7 · see citing works
Generative Artificial Intelligence and the Media: A Bibliometric Analysis in Academic Research
Hacek, J. (2025) · Media Literacy and Academic Research · doi:10.34135/mlar-25-01-02 · Cited by 4 · see citing works
Fârte, G.-I., Obadă, D.-R., Babii, A.-N., & Dabija, D. (2025) · Journal of Research in Interactive Marketing · doi:10.1108/jrim-10-2024-0473 · Cited by 3 · see citing works
Elishar-Malka, V., & Ariel, Y. (2026) · Journalism and Media · doi:10.3390/journalmedia7020116 · Cited by 2 · see citing works
Andreadis, K., Buderer, N. M. F., & Langford, A. T. (2025) · JMIR Formative Research · doi:10.2196/71971 · Cited by 1 · see citing works
From Framework to Nodes: Investigating Semantic Interaction in Generative AI Dialogue Flows
Ma, Y., Li, K., & Liu, Y. (2026) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2026.2629001 · Cited by 1 · see citing works
Theodorakopoulos, L., & Theodoropoulou, A. (2026) · Electronics · doi:10.3390/electronics15071405 · Cited by 1 · see citing works
ÓhÉigeartaigh, S. S., & Corsi, G. (2026) · Philosophy & Technology · doi:10.1007/s13347-026-01095-4 · Cited by 1 · see citing works
Self-citationBoatwright, B., DiRusso, C., & Pyle, A. S. (2026) · Public Relations Review · doi:10.1016/j.pubrev.2026.102716 · Cited by 1 · see citing works
Ye, H. (2026) · Social Media + Society · doi:10.1177/20563051261452130 · Cited by 1 · see citing works
Chani, T., & Olugbara, O. O. (2025) · Journalism and Media · doi:10.3390/journalmedia6040205 · Cited by 1 · see citing works
Shin, D. (2026) · Oxford University Press eBooks · doi:10.1093/9780197849644.003.0004 · Cited by 1 · see citing works
Sarmiento, H., Ortiz, J., Bravo-Márquez, F., dos Santos, M. L. B., & Valenzuela, S. (2024) · IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology (WI-IAT) · doi:10.1109/wi-iat62293.2024.00017
Embracing the Future of AI in Strategic Communication
Cheng, Y., He, W., Qiao, J., & Verčič, D. (2025) · doi:10.1002/9781394275434.oth
Morandín-Ahuerma, F. (2025) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2025.2565393
Liu, L., Huang, B., & Lai, S. (2025) · Scientific Reports · doi:10.1038/s41598-025-24746-6
ViralBERT-ID: Transformers-based Indonesian Tweet Virality Prediction Model
Krisnanik, E., Putri, D. S., Muhaimin, A., Umam, F., Haniefardy, A., & Diyasa, I. G. S. M. (2025) · International Conference on Informatics, Multimedia, Cyber and Information System (ICIMCIS) · doi:10.1109/icimcis68501.2025.11327213
Dongqi, Y., Ren, W., Junchen, Y., Yuduo, W., He, Y., & Zhang, H. (2026) · Journal of Cleaner Production · doi:10.1016/j.jclepro.2026.147591
Pasquini, D., Vocca, P., & Amati, G. (2025) · International Conference on Social Networks Analysis, Management and Security (SNAMS) · doi:10.1109/snams67467.2025.11391134
Francis, T. (2026) · Roczniki Nauk Społecznych · doi:10.18290/rns2026.0007
Social Media and Artificial Intelligence: Information, Bias, and Trust
Santos, S., Santos, J. D., & Pires, P. B. (2026) · Springer eBooks · doi:10.1007/978-3-031-75316-9_96-1
Dzingel, H., Meyer, A., Richter, E., Carpenter, J. P., & Richter, D. (2026) · Teaching and Teacher Education · doi:10.1016/j.tate.2026.105498
From Antisemitic Conspiracies to Islamophobic Echoes
Ali, S., Akhtar, M., & Akhtar, S. (2026) · IGI Global eBooks · doi:10.4018/979-8-3373-6068-3.ch003
Griffiths, L., & Ismagilova, E. (2026) · IFIP advances in information and communication technology · doi:10.1007/978-3-032-16775-0_8
Evaluating Explainable AI Implementation and User Agency Across Major Social Media Platforms
Alam, S., Pawade, A., Ayub, M. S., Rehman, S. U., & Ayub, A. (2025) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata66926.2025.11481468
Afghanistan’s Digital Diaspora: Conflicted Constructions of Nation and Identity
Mehran, W., & Drzewiecka, J. A. (2026) · Social Media + Society · doi:10.1177/20563051261434800
AI in hybrid conflicts: towards a legal framework for cognitive warfare
Manitra, R. R. M., & Prabandari, A. P. (2026) · Cogent Social Sciences · doi:10.1080/23311886.2026.2681304
Shankar, R., Lim, A., & Xu, Q. (2026) · Journal of dementia and Alzheimer's disease · doi:10.3390/jdad3030034
Digital and Substance Dependence in the Post-Digital Era
Romeo, V. M. (2026) · Encyclopedia · doi:10.3390/encyclopedia6070160
Naranjo-Vinueza, A., Martín, S. C., González, M. C., & Moreno, P. N.-B. (2026) · Frontiers in Political Science · doi:10.3389/fpos.2026.1850023
Che, F., Qin, Q., & Guo, L. (2026) · International Political Science Review · doi:10.1177/01925121261475523
Boratyn, D. (2026) · Crime Prevention and Community Safety · doi:10.1057/s41300-026-00251-4
Qureshi, K. A., Sabih, M., & Siddiqui, M. S. (2026) · Future Internet · doi:10.3390/fi18100563
- Authors
- Corsi, G.
- Affiliations
- University of Cambridge (GB)
- Corresponding author
- Corsi, Giulio
- ORCID
- Corsi, Giulio
- Venue
- EPJ Data Science
- Published
- 2024-03-07
- Volume
- 13
- Issue
- 1
- ISSN
- 2193-1127 (ISSN-L)
- OpenAlex
- W4392563408
- Cited by
- 45
- References
- 28
- Citations by year
- 2026: 22 · 2025: 21 · 2024: 2
- Type
- dissertation
- Language
- en
- Open access
- gold
- doi.org
- Locations
- EPJ Data Science (publishedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Topics
- Social Sciences › Sociology and Political Science › Misinformation and Its Impacts
- Physical Sciences › Physics and Astronomy › Statistical and Nonlinear Physics › Complex Network Analysis Techniques
- Physical Sciences › Computer Science › Information Systems › Spam and Phishing Detection
- Keywords
- algorithmic amplification, Twitter, recommender systems, disinformation, misinformation, engagement, toxicity, political bias, verified accounts, COVID-19 pandemic, climate change, content visibility
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
