Handling and Presenting Harmful Text in NLP Research
Open accessThe DOI matches its Crossref record; 17 of the 17 references it cites were found in a registry or an index.
Abstract
Text data can pose a risk of harm.However, the risks are not fully understood, and how to handle, present, and discuss harmful text in a safe way remains an unresolved issue in the NLP community.We provide an analytical framework categorising harms on three axes: (1) the harm type (e.g., misinformation, hate speech or racial stereotypes); (2) whether a harm is sought as a feature of the research design if explicitly studying harmful content (e.g., training a hate speech classifier), versus unsought if harmful content is encountered when working on unrelated problems (e.g., language generation or part-of-speech tagging); and (3) who it affects, from people (mis)represented in the data to those handling the data and those publishing on the data.We provide advice for practitioners, with concrete steps for mitigating harm in research and in publication.To assist implementation we introduce HARMCHECK -a documentation standard for handling and presenting harmful text in research.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Language Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372
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 · cited 1×
“Harms in datasets can transfer to the models trained on them (Bender et al., 2021; Rauh et al., 2022).”
BackgroundOn the Opportunities and Risks of Foundation Models
Bommasani, R., Hudson, D. A., Adeli, E., Altman, R., Arora, S., von Arx, S., Bernstein, M. S., Bohg, J., Bosselut, A., Brunskill, E., et al. (2021) · arXiv · doi:10.48550/arxiv.2108.07258 · Cited by 2,676
Ahmed, S. (2007) · Feminist Theory · doi:10.1177/1464700107078139 · Cited by 2,144 · cited 1×
“Representational harms, for example, emerge from sexist, racist, ableist, and otherwise unjust historical, cultural and societal norms, which are then embedded in data (Blodgett et al., 2020; Ahmed, 2007).”
BackgroundPrior exposure increases perceived accuracy of fake news
Pennycook, G., Cannon, T. D., & Rand, D. G. (2018) · Journal of Experimental Psychology General · doi:10.1037/xge0000465 · Cited by 1,268 · cited 1×
“For instance, exposure to false headlines increases the chance of their claims being accepted and normalised, even when the reader knows they are false (Pennycook et al., 2018).”
BackgroundBender, E. M., & Friedman, B. (2018) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00041 · Cited by 854 · cited 2×
“Each section is designed to be filled in as a statement (such as in a data statement (Bender and Friedman, 2018) or model card (Mitchell et al.”
Background“Everyone wants to do the model work, not the data work”: Data Cascades in High-Stakes AI
Sambasivan, N., Kapania, S., Highfill, H., Akrong, D., Paritosh, P., & Aroyo, L. M. (2021) · Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3411764.3445518 · Cited by 781 · cited 1×
“Text data is fundamental to nearly all areas of language processing, and good quality data is critical for better performing, safer and more robust models (Sambasivan et al., 2021; Xu et al., 2021).”
BackgroundAlgorithmic injustice: A relational ethics approach
Birhane, A. (2021) · Patterns · doi:10.1016/j.patter.2021.100205 · Cited by 707 · cited 2×
“Harms arise from ‘hyper-visibility’ (Noble, 2013), where excessive data portraying negative attributes such as stereotypes is collected (Birhane and Prabhu, 2021).”
BackgroundSelf-citationRealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models
Gehman, S., Gururangan, S., Sap, M., Choi, Y., & Smith, N. A. (2020) · Findings of the Association for Computational Linguistics: EMNLP 2020 · doi:10.18653/v1/2020.findings-emnlp.301 · Cited by 693
Ethical and social risks of harm from Language Models
Weidinger, L., Mellor, J., Rauh, M., Griffin, C., Uesato, J., Huang, P.-S., Cheng, M., Glaese, M., Balle, B., Kasirzadeh, A., et al. (2021) · arXiv · doi:10.48550/arxiv.2112.04359 · Cited by 477
Self-citation
Cited by
Auditing large language models: a three-layered approach
Mökander, J., Schuett, J., Kirk, H. R., & Floridi, L. (2024) · AI and Ethics · doi:10.1007/s43681-023-00289-2 · Cited by 289
Self-citationThe life cycle of large language models in education: A framework for understanding sources of bias
Lee, J., Hicke, Y., Yu, R., Brooks, C., & Kizilcec, R. F. (2024) · British Journal of Educational Technology · doi:10.1111/bjet.13505 · Cited by 119
“…hate speech, stereotypes), whether harmful content is sought out for the specific application (eg, to learn how to identify it better going forward) or not, and who is affected by the harmful content (eg, individuals represented in the dataset, demographic groups) (Kirk et al., 2022).”
Background“They only care to show us the wheelchair”: disability representation in text-to-image AI models
Mack, K. A., Qadri, R., Denton, R., Kane, S. K., & Bennett, C. L. (2024) · Proceedings of the CHI Conference on Human Factors in Computing Systems · doi:10.1145/3613904.3642166 · Cited by 67
Data-Efficient Strategies for Expanding Hate Speech Detection into Under-Resourced Languages
Röttger, P., Nozza, D., Bianchi, F., & Hovy, D. (2022) · Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/2022.emnlp-main.383 · Cited by 18
Metadata
- Authors
- Kirk, H., Birhane, A., Vidgen, B., & Derczynski, L.
- Venue
- Findings of the Association for Computational Linguistics: EMNLP 2022
- Published
- 2022
- Pages
- 497-510
- OpenAlex
- W4385573644
- Cited by
- 25
- References
- 17
- FWCI
- 6.16
- Type
- conference paper
- Language
- en
- Open access
- gold
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Handling and Presenting Harmful Text in NLP Research. https://app.cochat.ai/papers/10.18653/v1/2022.findings-emnlp.35
- The paper
- Kirk, H., Birhane, A., Vidgen, B., & Derczynski, L. (2022). Handling and Presenting Harmful Text in NLP Research. Findings of the Association for Computational Linguistics: EMNLP 2022. https://doi.org/10.18653/v1/2022.findings-emnlp.35
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.