The Role of Expertise in Effectively Moderating Harmful Social Media Content
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?
Prior research has shown that so-called 'ground truth' labels are shaped by specific contexts, such as annotators' backgrounds and expertise, as well as broader organizational control.
“Prior research has shown that so-called 'ground truth' labels are shaped by specific contexts, such as annotators' backgrounds and expertise, as well as broader organizational control [1, 76].”
From Disability-First AI Dataset Annotation: Co-designing Stuttered Speech Annotation Guidelines with People Who Stutter (2026) · cites this paper as [1]
Other research has uncovered the ambiguity and uncertainty inherent in socially constructed labels, such as gender, accent, and content toxicity.
“Other research has uncovered the ambiguity and uncertainty inherent in socially constructed labels, such as gender [85], accent [81], and content toxicity [1, 16].”
From Disability-First AI Dataset Annotation: Co-designing Stuttered Speech Annotation Guidelines with People Who Stutter (2026) · cites this paper as [1]
Recent studies and reporting have documented low wages, job insecurity, lack of healthcare benefits, and psychological toll for content moderators in AI data work.
“Recent studies and reporting on AI data work have documented low wages [90, 104], job insecurity [109], lack of healthcare benefits, and, for content moderators, the psychological toll of viewing disturbing content [1, 122].”
From 'The plan is just survival': Data Work in Kenya and the Regime of Entrapment (2026) · cites this paper as [1]
Cites
The Psychological Well-Being of Content Moderators
Steiger, M., Bharucha, T. J., Venkatagiri, S., Riedl, M. J., & Lease, M. (2021) · Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3411764.3445092 · Cited by 168 · see citing works
Addressing age-related bias in sentiment analysis,
Diaz, M., Johnson, I., Lazar, A., Piper, A. M., & Gergle, D. (2018) · Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3173574.3173986 · Cited by 138 · cited 1× · see citing works
“Another well-studied source of disagreement is annotators’ subjectivity, as their race [58, 61], age [24, 66], education level, personality [37, 66], language [66], personal beliefs and values [16, 68], political orientation [37], and knowledge of hate speech [84] can affect annotation practices…”
Contestability For Content Moderation
Vaccaro, K., Xiao, Z., Hamilton, K., & Karahalios, K. (2021) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3476059 · Cited by 77 · see citing works
The poverty of ethical AI: impact sourcing and AI supply chains
Muldoon, J., Cant, C., Graham, M., & Ustek Spilda, F. (2025) · AI & SOCIETY · doi:10.1007/s00146-023-01824-9 · Cited by 70 · see citing works
The psychological impacts of content moderation on content moderators: A qualitative study
Spence, R., Bifulco, A., Bradbury, P., Martellozzo, E., & DeMarco, J. (2023) · Cyberpsychology: Journal of Psychosocial Research on Cyberspace · doi:10.5817/cp2023-4-8 · Cited by 48 · cited 1× · see citing works
“Table 1 summarizes the relevant backgrounds our annotators have, in addition to their linguistic knowledge.”
Content Moderator Mental Health, Secondary Trauma, and Well-being: A Cross-Sectional Study
Spence, R., Bifulco, A., Bradbury, P., Martellozzo, E., & DeMarco, J. (2024) · Cyberpsychology, Behavior, and Social Networking · doi:10.1089/cyber.2023.0298 · Cited by 22 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
Tang, X., Li, J., & Wu, S. (2026) · Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3772318.3790405 · Cited by 4 · see citing works
'The plan is just survival': Data Work in Kenya and the Regime of Entrapment
Kapania, S., Yang, T., Abdelkadir, N. A., Scheuerman, M. K., Miceli, M., Taylor, A. S., & Fox, S. E. (2026) · Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3772318.3791097 · Cited by 2 · see citing works
“Recent studies and reporting on AI data work have documented low wages [90, 104], job insecurity [109], lack of healthcare benefits, and, for content moderators, the psychological toll of viewing disturbing content [1, 122].”
Self-citation
- Authors
- Abdelkadir, N. A., Yang, T., Kapania, S., Estefanos, M., Gebrekidan, F. B., Zelalem, Z., Ali, M., Berhe, R., Baker, D., Talat, Z., Miceli, M., Hanna, A., & Gebru, T.
- Affiliations
- University of Minnesota (US)
- Weizenbaum Institute (DE)
- Carnegie Mellon University (US)
- DAIR Institute (US)
- Independent Researcher, Nairobi, Kenya
- Freelance Journalist, Montreal, Quebec, Canada
- Independent Researcher, Stockholm, Sweden
- Independent Researcher, New York, New York, USA
- University of Edinburgh (GB)
- ORCID
- Venue
- Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems
- Publisher
- ACM
- Published
- 2025-04-25
- Pages
- 1-21
- OpenAlex
- W4409748992
- Cited by
- 5
- References
- 6
- Citations by year
- 2026: 5
- FWCI
- 4.74
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- CHI Conference on Human Factors in Computing Systems (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Hate Speech and Cyberbullying Detection
- Social Sciences › Law › Freedom of Expression and Defamation
- Social Sciences › Sociology and Political Science › Misinformation and Its Impacts
- Keywords
- social media, expertise, content moderation, online harms
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
