Correction to: Excavating AI: the politics of images in machine learning training sets

Crawford, Kate; Paglen, Trevor

doi:10.1007/s00146-021-01301-1

Cited by 153Ethics and Social Impacts of AILaw, AI, and Intellectual PropertyDigitalization, Law, and Regulation

Correction to: Excavating AI: the politics of images in machine learning training sets

Open access

Crawford, Kate, Paglen, Trevor·2021·AI & SOCIETY

DOI 10.1007/s00146-021-01301-1

Open this paper in CoChat — free account Open PDF

Integrity signals

Registry

Crossref · 10.1007/s00146-021-01301-1

Cites

3 of 3 references verified0 citation errors · 0 not checkable

Statistics

Recomputing p-values needs the paper’s full text.

Claims the paper makes

The text we found for this paper looks like another document, so its claims and self-check weren’t run. The registry and citation checks still apply.

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The text we found for this paper looks like another document, so the self-check wasn’t run.

ImageNet: A large-scale hierarchical image database
Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206848 · Cited by 65,333 · see citing works
Coding facial expressions with Gabor wavelets
Lyons, M. J., Akamatsu, S., Kamachi, M., & Gyoba, J. (2002) · International Conference on Automatic Face and Gesture Recognition · doi:10.1109/afgr.1998.670949 · Cited by 2,214 · see citing works
Diversity in Faces.Preprint
Merler, M., Ratha, N., Feris, R., & Smith, J. R. (2019) · arXiv (Cornell University) · Cited by 111 · see citing works

Flagged first, then by citations

Sorted by citations

Fairness and Bias in Artificial Intelligence: A Brief Survey of Sources, Impacts, and Mitigation Strategies
Ferrara, E. (2023) · Sci · doi:10.3390/sci6010003 · Cited by 875 · see citing works
“To address this issue, some companies have implemented “gender decoder” tools that analyze job postings and suggest changes to reduce gender bias [17,37].”
Background
Artificial intelligence governance: Ethical considerations and implications for social responsibility
Camilleri, M. A. (2024) · Expert Systems · doi:10.1111/exsy.13406 · Cited by 331 · see citing works
“Their dialogue formats enable them to respond to questions (Thorp, 2023), to admit their mistakes (Barrot, 2023), and to even reject requests (Crawford & Paglen, 2021), if they are not recognized as appropriate.”
Background
Potential Risks of Artificial Intelligence Integration into School Education: A Systematic Review
Karan, B., & Angadi, G. R. (2023) · Bulletin of Science Technology & Society · doi:10.1177/02704676231224705 · Cited by 108 · see citing works
“For instance, students of different color in K-12 schools are subjected to the discrimination at entry for incorrect identi ﬁ cation by AI (Crawford & Paglen, 2021).”
Background
Whose AI Dream? In search of the aspiration in data annotation.
Wang, D., Prabhat, S., & Sambasivan, N. (2022) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3491102.3502121 · Cited by 70 · see citing works
Web scraping for research: Legal, ethical, institutional, and scientific considerations
Brown, M. A., Gruen, A., Maldoff, G., Messing, S., Sanderson, Z., & Zimmer, M. T. (2025) · Big Data & Society · doi:10.1177/20539517251381686 · Cited by 48 · see citing works
“…are increasingly fed into machine learning models and other computational tools, they may be repur-posed in contexts far removed from the original research intent, such as facial recognition, predictive policing, or behavioral pro ﬁ ling (Crawford and Paglen, 2021; Scheuerman et al., 2021).”
Artificial Aesthetics and Ethical Ambiguity: Exploring Business Ethics in the Context of AI-driven Creativity
Xu, C. G., Sun, Y., & Zhou, H. (2024) · Journal of Business Ethics · doi:10.1007/s10551-024-05837-2 · Cited by 40 · see citing works
“These considerations include questions about authorship, creative control, and the potential impact of AI-generated content on human artistic expression (Crawford & Paglen, 2021).”
A Synthesis of AI in higher education: Shaping the future
Jacques, P. H., Moss, H. K., & Garger, J. (2024) · Journal of Behavioral and Applied Management · doi:10.21818/001c.122146 · Cited by 43 · see citing works
“AI algorithms thus perpetuate and amplify biases in education, affecting student evaluations, admissions decisions, and learning opportunities (Akgun & Greenhow, 2021; Crawford & Paglen, 2021).”
Background
From rules to examples: Machine learning's type of authority
Campolo, A., & Schwerzmann, K. (2023) · Big Data & Society · doi:10.1177/20539517231188725 · Cited by 37 · see citing works
“…of precarious labor on platforms like Amazon Mechanical Turk; dataset production is also reproduction of a preexisting onto-epistemology (Crawford and Paglen, 2021: 1113) or ground truth (Jaton, 2017; 2021): here, the hierarchical clas-si ﬁ catory schema derived from the lexical…”
Background
ECCDN-Net: A deep learning-based technique for efficient organic and recyclable waste classification
Islam, M. S. B., Sumon, M. S. I., Majid, M. E., Kashem, S. B. A., Nashbat, M., Ashraf, A., Khandakar, A. A., Kunju, A. K. A., Hasan-Zia, M., & Chowdhury, M. E. H. (2024) · Waste Management · doi:10.1016/j.wasman.2024.12.023 · Cited by 31 · see citing works
Creative data justice: a decolonial and indigenous framework to assess creativity and artificial intelligence
Arora, P. (2024) · Information Communication & Society · doi:10.1080/1369118x.2024.2420041 · Cited by 35 · see citing works
“An ‘arche-ology of datasets’ (Crawford & Paglen, 2021) approach can unpack the structural components of training data – images and labels, and ideologies that shape data classifications.”
The generic uniqueness of AI imagery: A critical approach to Dall-E as semiotic technology
Westberg, G., & Kvåle, G. (2024) · Discourse & Society · doi:10.1177/09579265241274788 · Cited by 26 · see citing works
“These lines of social issues did not result from the recent advent of AI images, but for decades have been part of the training sets used for machine learning, such as ImageNet (Crawford and Paglen, 2021).”
Adaptive Epistemology: Embracing Generative AI as a Paradigm Shift in Social Science
Punziano, G. (2025) · Societies · doi:10.3390/soc15070205 · Cited by 28 · see citing works
Cross-cultural approaches to creative media content in the age of AI
Chateau, L., Arora, P., & Herman, L. M. (2025) · Media Culture & Society · doi:10.1177/01634437251328188 · Cited by 21 · see citing works
“Templates seem to embody a prescriptive form that can be filled in using stock images and other material provided by the host platforms, where skewed datasets and biased data can again be a problem (Crawford and Paglen, 2021).”
Unpacking artificial intelligence in elementary education: A comprehensive thematic analysis systematic review
Boulhrir, T., & Hamash, M. (2025) · Computers and Education Artificial Intelligence · doi:10.1016/j.caeai.2025.100442 · Cited by 19 · see citing works
Continual Forgetting for Pre-Trained Vision Models
Zhao, H., Ni, B., Fan, J., Wang, Y., Chen, Y., Meng, G., & Zhang, Z. (2024) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52733.2024.02705 · Cited by 23 · see citing works
What is research data “misuse”? And how can it be prevented or mitigated?
Pasquetto, I. V., Cullen, Z., Thomer, A. K., & Wofford, M. F. (2024) · Journal of the Association for Information Science and Technology · doi:10.1002/asi.24944 · Cited by 17 · see citing works
“There are plenty of documented examples of this kind of misuse, such as the use of racially biased data in legal and police procedures (Green, 2021), and the use of data containing offensive language and images in datasets used to train image processing algorithms (Crawford & Paglen, 2021).”
Imaginaries of democratization and the value of open environmental data: Analysis of Microsoft's planetary computer
Lukacz, P. M. (2024) · Big Data & Society · doi:10.1177/20539517241242448 · Cited by 15 · see citing works
“As Crawford and Paglen (2021) recount, in the early 2010s, the ImageNet dataset became a “ critical asset for computer-vision research.”
The Digital Faces of Oppression and Domination: A Relational and Egalitarian Perspective on the Data-driven Society and its Regulation
Naudts, L. (2024) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3630106.3658934 · Cited by 15 · see citing works
Situating the social issues of image generation models in the model life cycle: a sociotechnical approach
Katirai, A., García, N., Ide, K., Nakashima, Y., & Kishimoto, A. (2024) · AI and Ethics · doi:10.1007/s43681-024-00517-3 · Cited by 17 · see citing works
“Finally, as Bianchi et al. [82] note, issues of bias are not easily mitigated, given that any understanding of an image is inherently subjective [17].”
Understanding and evaluating harms of AI-generated image captions in political images
Sarhan, H., & Hegelich, S. (2023) · Frontiers in Political Science · doi:10.3389/fpos.2023.1245684 · Cited by 14 · see citing works
AI depictions of psychiatric diagnoses: a preliminary study of generative image outputs in Midjourney V.6 and DALL-E 3
Flathers, M., Smith, G., Wagner, E. S., Fisher, C. E., & Torous, J. B. (2024) · BMJ Mental Health · doi:10.1136/bmjment-2024-301298 · Cited by 14 · see citing works
“15 Second, as these data sets scale, they must source content from less curated, lower-quality sections of the internet, expanding into more chaotic, unmoderated domains.”
Rethinking model prototyping through the MedMNIST+ dataset collection
Doerrich, S., Salvo, F. D., Brockmann, J., & Ledig, C. (2025) · Scientific Reports · doi:10.1038/s41598-025-92156-9 · Cited by 14 · see citing works
Copying style, Extracting value: Illustrators' Perception of AI Style Transfer and its Impact on Creative Labor
Porquet, J., Wang, S., & Chilton, L. B. (2025) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3706598.3713854 · Cited by 14 · see citing works
Assessing deep learning: a work program for the humanities in the age of artificial intelligence
Segessenmann, J., Stadelmann, T., Davison, A., & Dürr, O. M. (2023) · AI and Ethics · doi:10.1007/s43681-023-00408-z · Cited by 13 · see citing works
Leveraging Knowledge Graphs for AI System Auditing and Transparency
Waltersdorfer, L., & Sabou, M. (2024) · Journal of Web Semantics · doi:10.1016/j.websem.2024.100849 · Cited by 20 · see citing works
AI-Competent Individuals and Laypeople Tend to Oppose Facial Analysis AI
Ullstein, C., Engelmann, S., Papakyriakopoulos, O., Hohendanner, M., & Großklags, J. (2022) · ACM Conference on Equity and Access in Algorithms, Mechanisms, and Optimization (EAAMO) · doi:10.1145/3551624.3555294 · Cited by 12 · see citing works
Encoding/decoding in artificial intelligence: Global AI and local languages
Yin, F. (2025) · Media Culture & Society · doi:10.1177/01634437251360381 · Cited by 11 · see citing works
“Biases can also be generated from training datasets – the core source for machines to classify, model, and predict society and set epistemic parameters for how AI systems operate (Crawford and Paglen, 2021).”
Artificial Intelligence and the ethics of navigating ambiguity
Bennett, S. (2025) · Big Data & Society · doi:10.1177/20539517251347594 · Cited by 10 · see citing works
“Crawford and Paglen (2021) examined the politics of ‘ objectivity ’ in computer vision systems, charting how this is negotiated across multiple layers that are often hidden from view.”
When can we Kick (Some) Humans “Out of the Loop”? An Examination of the use of AI in Medical Imaging for Lumbar Spinal Stenosis
Muyskens, K., Ma, Y., Menikoff, J., Hallinan, J. T. P. D., & Savulescu, J. (2024) · Asian Bioethics Review · doi:10.1007/s41649-024-00290-9 · Cited by 10 · see citing works
Engineering equity: designing diversity-aware AI to reflect humanity
Shin, D. (2025) · AI & Society · doi:10.1007/s00146-025-02314-w · Cited by 10 · see citing works
“Transparency, which ensures that AI decisions are explainable and understandable, is a separate concern from diversity, as an AI system can be diverse but still operate as a black box (Crawford and Paglen 2021).”
Influential
Data, Annotation, and Meaning-Making: The Politics of Categorization in Annotating a Dataset of Faith-based Communal Violence
Rifat, M. R., Safir, A. H., Saha, S., Junaed, J. A., Saleki, M., Amin, M. R., & Ahmed, S. I. (2024) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3630106.3659030 · Cited by 9 · see citing works
The Pillars of AI Ethics
Ramya, R., Priya, S. B., Thamizhikkavi, P., & Anand, M. (2024) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3693-9173-0.ch004 · Cited by 10 · see citing works
Transhumanism as Capitalist Continuity: Branded Bodies in the Age of Platform Sovereignty
Lockhart, E. N. S. (2025) · Humans · doi:10.3390/humans5030021 · Cited by 9 · see citing works
“The human is no longer a sovereign individual but a calibrated entity molded to fit neoliberal imperatives, with profound ethical and political implications for autonomy and justice.”
Human Participants in AI Research: Ethics and Transparency in Practice
McKee, K. R. (2024) · IEEE Transactions on Technology and Society · doi:10.1109/tts.2024.3446183 · Cited by 12 · see citing works
Synthetic imaginaries of “sensitive” AI: On ambient amplification and jail(break)ing as method
Pilipets, E., & Geboers, M. (2025) · Platforms & Society · doi:10.1177/29768624251378110 · Cited by 9 · see citing works
Gender and the invisibility of care on Wikipedia
Ford, H., Pietsch, T., & Tall, K. (2023) · Big Data & Society · doi:10.1177/20539517231210276 · Cited by 7 · see citing works
“Despite the fact that adding female biog-raphies is often met with signi ﬁ cant hostility (Tripodi, 2021), ‘ there is no easy technical “ ﬁ x ” [to bias problems] by shifting demographics, deleting offensive terms or seeking equal representation by skin tone ’ (Crawford and Paglen, 2021: 1113).”
Background
Beyond Fairness in Computer Vision: A Holistic Approach to Mitigating Harms and Fostering Community-Rooted Computer Vision Research
Gebru, T., & Denton, E. (2024) · Foundations and Trends® in Computer Graphics and Vision · doi:10.1561/0600000102 · Cited by 7 · see citing works
Machine's Eye View: Postmodern Data Science and the Politics of Ground Truth
Sadowski, J. (2025) · Science Technology & Human Values · doi:10.1177/01622439251331138 · Cited by 7 · see citing works
“Multiple studies have now laid out, in painstaking detail, how ImageNet is rife with racist, sexist, and homophobic labels on images, including out-right slurs, alongside categories that label images with bizarre moral judgements, such as: “ Bad Person, Call Girl, Drug Addict, Closet Queen, Convict, Crazy, Failure, Fl…”
Labour And Exploitation Processes in Artificial Intelligence: Example of Digital Taylorism in Data Labelling
YILMAZ, Ö., & Bostancı, M. (2025) · Toplum ve Kültür Araştırmaları Dergisi · doi:10.48131/jscs.1646235 · Cited by 7 · see citing works
Challenges and Responsibilities in the UX Design of Text-to-Image AI Models: A Discussion Through a Comparative Heuristics Evaluation
Vacanti, A., Burlando, F., Paz, I., & Menichinelli, M. (2024) · Temes de disseny · doi:10.46467/tdd40.2024.156-175 · Cited by 7 · see citing works
Composing the Assemblage: Probing Aesthetic and Technical Dimensions of Artistic Creation with Machine Learning
Gioti, A.-M., Einbond, A., & Born, G. (2022) · Computer Music Journal · doi:10.1162/comj_a_00658 · Cited by 7 · see citing works
Memory of the multitude and representation in AI-generated images of war
Laba, N., Roman, N., & Parmelee, J. H. (2025) · Memory Mind & Media · doi:10.1017/mem.2025.10011 · Cited by 6 · see citing works
“…sources, to learn patterns and produce visuals, reflect the cultural, social, and historical prejudices embedded in their original context of use (Crawford and Paglen 2021), including ideological positioning of news outlets as well as those stemming from civic reporting (Maier-Rothe et al. 2014).”
“We Are Bulletproof”: The Transcultural Power of Fandom in # StopAsianHate
Zhang, A., & ong, C. (2024) · Sociological Inquiry · doi:10.1111/soin.12590 · Cited by 6 · see citing works
AI and the Evolution of Artistic Expression
Yadav, M., Kumar, M., Sahoo, A., & Rathnasiri, M. S. H. (2024) · Advances in media, entertainment and the arts (AMEA) book series · doi:10.4018/979-8-3693-3916-9.ch002 · Cited by 6 · see citing works
An empirical study of ChatGPT-related projects and their issues on GitHub
Lin, Z., Zhang, N., Liu, C. C., & Zheng, Z. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.126113 · Cited by 6 · see citing works
“After receiving natural language input, chatbots can generate responses in various forms such as text, code, images, and video [8, 12, 13, 33].”
Advancing a Material and Epistemological Turn in the Study of AI: A Review and New Directions for Journalism Research
Schjøtt, A. (2025) · Digital Journalism · doi:10.1080/21670811.2025.2485255 · Cited by 7 · see citing works
“Similarly, crawford and Paglen (2021) interrogate the underlying assumptions of several datasets for visual Ai systems by exploring their taxonomies and how their construction impacts the workings of these systems in societies.”
Repairing what’s not broken – Algorithm repair manual as reflexivity device
Stefanija, A. P., & Pierson, J. (2024) · Convergence The International Journal of Research into New Media Technologies · doi:10.1177/13548565241261978 · Cited by 5 · see citing works
“Data used both for training the algorithms and as computational input for algorithmic outputs is often labeled based on essentialist choices, prioritizing one meaning of the data over other (Crawford and Paglen, 2019, 2021).”
Investigating AI systems: examining data and algorithmic bias through hermeneutic reverse engineering
Shukla, N. (2025) · Frontiers in Communication · doi:10.3389/fcomm.2025.1380252 · Cited by 5 · see citing works
Emotion AI in Job Interviews: Injustice, Emotional Labor, Identity, and Privacy
Ingber, A. S., & Andalibi, N. (2025) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3715275.3732002 · Cited by 5 · see citing works
Reducing AI Model Biases with a Bilevel Learning Framework: A Case Study of Leveraging Twitter Data for Damage Estimation
Bai, W., Ye, X., Xie, Y., Zandt, S. V., Huang, X., & Sengupta, D. (2025) · Annals of the American Association of Geographers · doi:10.1080/24694452.2025.2536182 · Cited by 4 · see citing works
Geographies of Environmental Data 1
Ritts, M. (2025) · Progress in Environmental Geography · doi:10.1177/27539687251398578 · Cited by 4 · see citing works
“The vivid renderings of Kate Crawford and Vladen Joler ’ s “ Atlas of AI, ” (2018) for instance, distill complex data geographies alongside questions of labor, extraction, and material histories of power/knowledge (See also: Crawford and Paglen 2021).”
Algorithmic gaze and subject occlusion: a medical ethical critique of artificial intelligence diagnosis and treatment from a foucaultian perspective
Dai, Y., & Guo, S. (2025) · Medicine Health Care and Philosophy · doi:10.1007/s11019-025-10318-y · Cited by 4 · see citing works
AI and big data in disaster response: Ethical and practical challenges
Wood, E. X. (2025) · Journal of dynamic disasters. · doi:10.1016/j.jdd.2025.100041 · Cited by 4 · see citing works
Fairness in machine learning: definition, testing, debugging, and application
Gao, X., Shen, C. Q., Jiang, W., Lin, C., Li, Q., Wang, Q., Li, Q., & Guan, X. (2024) · Science China Information Sciences · doi:10.1007/s11432-023-4060-x · Cited by 4 · see citing works
Spectacle, Surveillance, and the Ironies of Visual Politics in the Age of Autonomous Images
Reinhardt, M. (2023) · Political Theory · doi:10.1177/00905917231173442 · Cited by 5 · see citing works
“…machine learning training sets, drawn from the trillions of pictures populating the internet, rely on classifications and labels that often reproduce contemporary prejudices or reanimate older ones, giving new life to such discourses as phrenology and physiognomy (Crawford and Paglen 2021).”
Background
Confronting algorithmic management using subject access requests: Insights using the case of food deliveries
Perrig, L. (2023) · The Economic and Labour Relations Review · doi:10.1017/elr.2023.55 · Cited by 5 · see citing works
“The nature of the data however can be revealing of the outcome of algorithmic decision-making (Crawford and Paglen 2021).”
Background
The ripple effect of dataset reuse: Contextualising the data lifecycle for machine learning data sets and social impact
Park, J., & Cordell, R. (2023) · Journal of Information Science · doi:10.1177/01655515231212977 · Cited by 4 · see citing works
AI-generated imagery in ethnography: A reflective framework
Chen, J. (2025) · Big Data & Society · doi:10.1177/20539517251403957 · Cited by 4 · see citing works
“…imagery is not only consistent with ethnographic commitments but also essential for interrogating and balancing the new politics of representation, since “ the whole endeavor of collecting images, categorizing them, and labeling them is itself a form of politics ” (Crawford and Paglen, 2021: 1113).”
Strategic Framework for Implementing Artificial Intelligence in Organizations from Conceptualization to Practical Execution and Performance Evaluation
Gupta, A., Tyagi, L. K., Sisodia, V. S., & Verma, A. (2024) · International Conference on Sustainable Computing and Integrated Communication in Changing Landscape of AI (ICSCAI) · doi:10.1109/icscai61790.2024.10866620 · Cited by 3 · see citing works
Combining the Strengths of LLMs and Persuasive Technology to Combat Cyberhate
Almaliki, M., Almars, A. M., Aljuhani, K. O., & Atlam, E.-S. (2025) · Computers · doi:10.3390/computers14050173 · Cited by 3 · see citing works
The Moral Economy of AI
Chen, Y., Lindtner, S., & Sun, Y. (2025) · Decennial Aarhus Conference · doi:10.1145/3744169.3744173 · Cited by 3 · see citing works
Perceived gendered self-representation on Tinder using machine learning
Asadchy, Y., Karjus, A., Mukhina, K. D., & Schich, M. (2024) · Humanities and Social Sciences Communications · doi:10.1057/s41599-024-03801-z · Cited by 3 · see citing works
“Since we have gender-labeled data, we do not have to rely on biased annotated datasets like ImageNet and carry on these biases (Crawford and Paglen, 2021).”
Efficient Bias Mitigation Without Privileged Information
Zarlenga, M. E., Sankaranarayanan, S., Andrews, J. T. A., Shams, Z., Jamnik, M., & Xiang, A. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-73220-1_9 · Cited by 4 · see citing works
A framework for assessing reliability of observer annotations of aerial wildlife imagery, with insights for deep learning applications
Converse, R. L., Lippitt, C. D., Sesnie, S. E., Harris, G. M., Butler, M. J., & Stewart, D. R. (2025) · PLoS ONE · doi:10.1371/journal.pone.0316832 · Cited by 3 · see citing works
The Algorithmic Construction of Epistemic Injustice
McInerney, T. (2025) · Palgrave socio-legal studies · doi:10.1007/978-3-032-07581-9_5 · Cited by 4 · see citing works
Practical Continual Forgetting for Pre-Trained Vision Models
Zhao, H., Zhu, F., Ni, B., Zhu, F., Meng, G., & Zhang, Z. (2026) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2026.3654115 · Cited by 3 · see citing works
How to Make an Artificial Intelligence Algorithm “Ecological”? Insights from a holistic perspective
Mascio, T. D., Caruso, F., & Peretti, S. (2023) · Biannual Conference of the Italian SIGCHI Chapter (CHItaly) · doi:10.1145/3605390.3605398 · Cited by 3 · see citing works
Educating artificial intelligence following the child learning development trajectories
Peretti, S., Caruso, F., Valente, G., Pomante, L., & Mascio, T. D. (2025) · Behaviour and Information Technology · doi:10.1080/0144929x.2025.2455390 · Cited by 3 · see citing works
Mitigating Demographic Bias in ImageNet: A Comprehensive Analysis of Disparities and Fairness in Deep Learning Models
Gitonga, C. K., Murithi, D., & Too, E. C. (2025) · European Journal of Artificial Intelligence and Machine Learning · doi:10.24018/ejai.2025.4.2.51 · Cited by 3 · see citing works
Assessing Deep Learning: A Work Program for the Humanities in the Age of Artificial IntelligencePreprint
Segessenmann, J., Stadelmann, T., Davison, A., & Dürr, O. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4554234 · Cited by 3 · see citing works
When Being a Data Annotator Was Not Yet a Job: The Laboratory Origins of Dispersible Labor in Computer Vision Research
Li, Z. (2024) · Socius Sociological Research for a Dynamic World · doi:10.1177/23780231241259617 · Cited by 2 · see citing works
“…bias has offered important insights into the underrepresentation and stereotypical representation of minorities in machine learning datasets (Crawford and Paglen 2021; Denton et al. 2021; DeVries et al. 2019; Paullada et al. 2021; Shankar et al. 2017), this study calls for future research…”
Human-Centered AI (Also) for Humanistic Management
Facchini, A., & Mangili, F. (2024) · Humanism in business series · doi:10.1007/978-3-031-67155-5_11 · Cited by 4 · see citing works
Ethical Considerations and Implications of Generative AI in Computer Graphics
Routray, S. K. (2025) · IEEE Computer Graphics and Applications · doi:10.1109/mcg.2025.3570722 · Cited by 2 · see citing works
Breaking Bias: Addressing Ageism in Artificial Intelligence
Amundsen, D. (2025) · Journal of Ageing and Longevity · doi:10.3390/jal5030036 · Cited by 2 · see citing works
Responsibly Training Foundation Models: Actualizing Ethical Principles for Curating Large-Scale Training Datasets in the Era of Massive AI Models
Scheuerman, M. K., Zhao, D., Andrews, J. T. A., Birhane, A., Liao, Q. V., Panagiotidou, G., Chitre, P., Pine, K. H., Walker, S., Zhao, J., et al. (2025) · ACM Conference on Computer Supported Cooperative Work (CSCW) · doi:10.1145/3715070.3748285 · Cited by 2 · see citing works
Generative AI as a Philosophical Mirror: Machine Hallucination and the Aesthetics of Algorithmic Representation
Khokhar, M. S., Ayoub, M., & Zakria (2025) · Contemporary Visual Culture and Art · doi:10.63385/cvca.v1i1.98 · Cited by 2 · see citing works
Algorithmic HRM
Rismawati, R., & Salju, S. (2026) · IGI Global eBooks · doi:10.4018/979-8-3373-4307-5.ch001 · Cited by 2 · see citing works
INTRODUCTION: THE POLITICS AND PRACTICES OF COMPUTATIONAL SEEING
Hand, M., & Scarlett, A. (2023) · Photographies · doi:10.1080/17540763.2023.2189287 · Cited by 2 · see citing works
Discovering Explainability Requirements in ML-Based Software
Sporsem, T. (2024) · ACM/IEEE International Conference on Software Engineering: Companion Proceedings (ICSE-Companion) · doi:10.1145/3639478.3639807 · Cited by 2 · see citing works
Dialectics of training: A critique of recommendation engines’ aesthetic judgment
Musih, N., & Fisher, E. (2024) · Convergence The International Journal of Research into New Media Technologies · doi:10.1177/13548565241285568 · Cited by 2 · see citing works
“Indeed, it is an emic term, part of the technical discourse in the academic and professional ﬁ eld of data science (Crawford and Paglen, 2021).”
What Does It Mean to “Misuse” Research Data?
Pasquetto, I. V., Thomer, A., Acker, A., Chtena, N., & Desai, M. A. (2024) · Proceedings of the Association for Information Science and Technology · doi:10.1002/pra2.1108 · Cited by 2 · see citing works
Revolutionizing Digital Narratives: The Role of Semantic Web and Artificial Intelligence in StorytellingPreprint
Roy, A. (2025) · Preprints.org · doi:10.20944/preprints202503.1948.v3 · Cited by 2 · see citing works
Machine Learning Based Online English Teaching Monitoring Task System
Song, Y. (2023) · International Conference on Data Analytics, Computing and Artificial Intelligence (ICDACAI) · doi:10.1109/icdacai59742.2023.00085 · Cited by 1 · see citing works
Digital Curation and AI
Neudecker, C. (2023) · Edition Museum · doi:10.14361/9783839467107-013 · Cited by 1 · see citing works
Facial Recognition Technology in Context
Matulionytė, R., & Žalnieriūtė, M. (2024) · Cambridge University Press eBooks · doi:10.1017/9781009321211.002 · Cited by 1 · see citing works
The co-evolution of AI technology and information environment: Diagnosing social impacts and exploring governance strategies
Cha, S., Seo, B.-G., Kim, T., & Kim, J. (2024) · Journal of Infrastructure Policy and Development · doi:10.24294/jipd.v8i8.6605 · Cited by 1 · see citing works
Innovative Horizons
Sarao, P., Dar, A. A., Sindhuja, S., Malhotra, M., Jain, A., & Shruti (2025) · Advances in business strategy and competitive advantage book series · doi:10.4018/979-8-3693-4433-0.ch011 · Cited by 1 · see citing works
Queer ecological data: Where artificial intelligence meets the avian
Livio, M., & Sánchez-Querubín, N. (2025) · New Media & Society · doi:10.1177/14614448251313767 · Cited by 1 · see citing works
“Although the harms of birdsong data may be less immediately apparent, an example can be revealed by continuing the “archeology of datasets” initiated in the description of the Salvaging Birds project (Crawford and Paglen, 2021).”
Incorporation of Artificial Intelligence: Exploring Readiness Factors at School Level
Angadi, G. R., & Karan, B. (2026) · Bulletin of Science Technology & Society · doi:10.1177/02704676251407967 · Cited by 3 · see citing works
“Scholars (e.g., AI Shamsi et al., 2022; Crawford & Paglen, 2021; Karan & Angadi, 2023b; McStay, 2020) also worried that AI integration into K-12 education is associated with several signi ﬁ cant risks.”
Overload opacity: Countervisuality and glitch aesthetics in AI-Generated Brainrot on Instagram
Ungureanu, D. (2026) · Discourse Context & Media · doi:10.1016/j.dcm.2025.100972 · Cited by 1 · see citing works
Navigating the automation of the arts: Values, beliefs and emotions of arts practitioners using narrow AI
Frątczak, M., Ochu, E., Bates, J., & Medina-Perea, I. (2026) · Convergence The International Journal of Research into New Media Technologies · doi:10.1177/13548565261433690 · Cited by 1 · see citing works
“This includes CO 2 emissions arising from large-scale data collection and ML training models, as well as the use of minerals, e-waste, and their role in fossil fuel extraction (Brevini, 2020; Crawford and Paglen, 2021).”
Slicing the past to predict the future: Recasting data slicing as curatorial work in ML development and evaluation
Schjøtt, A. (2026) · Cambridge Forum on AI Culture and Society · doi:10.1017/cfc.2026.10014 · Cited by 1 · see citing works
“Research within Critical Dataset Studies has already drawn parallels between the production of ML datasets and archival practices (Crawford and Paglen 2021; Gebru et al. 2021; Jo and Gebru 2020; Pipkin 2020; Thylstrup 2022).”
Prompt Culture: From Lines to Surfaces in AI Image Generation
Musih, N. (2026) · Theory Culture & Society · doi:10.1177/02632764261455347 · Cited by 1 · see citing works
“Others adopt a critical stance, deliberately crafting prompts or influencing the dataset and the system in ways that expose or subvert generative AI’s limitations, making visible what escapes the conceptual nets (Crawford and Paglen, 2021; Paglen and Downey, 2024).”
Beyond ethics and aesthetics: perceived lay theory violations and agency disruption in AI-generated art
Dolese, M. J. (2026) · Frontiers in Psychology · doi:10.3389/fpsyg.2026.1716814 · Cited by 1 · see citing works
“While issues such as appropriation, job displacement, and copyright ambiguity remain central (Crawford and Paglen, 2021), they do not fully capture the range of public responses, which fluctuate between curiosity and fascination (Grassini and Koivisto, 2024; Shah, 2023) and persistent resistance…”
What If We Ran Our Own Show? GenAI and The Politics of Representation
Vanlee, F., & Baere, M. D. (2026) · Media and Communication · doi:10.17645/mac.12267 · Cited by 1 · see citing works
“This concerns both the visual materials themselves and the discursive labels with which the materials are rendered “legible” to genAI models (Crawford & Paglen, 2021).”
Generative AI Ethical Dimensions
Mahboubi, O. E., & Amnai, M. (2025) · IGI Global eBooks · doi:10.4018/979-8-3373-0365-9.ch019 · Cited by 1 · see citing works
Tool or medium? Expertise in training media and communications students in the Generative AI era
Atkinson, P., Dann, C., van der Nagel, E., Han, G.-S., Johnson, A., Rai, M., Richardson, L., & Stieven-Taylor, A. (2025) · Communication Research and Practice · doi:10.1080/22041451.2025.2582950 · Cited by 1 · see citing works
“Some of these issues are partially addressed in recent research on GenAI: Crawford and Paglen (2021, p. 1105) examine what narrow datasets in AI classifications mean for ‘race, gender, emotions, ability, sexuality, and personality’; Pavlik (2023, p. 92) urges journalism and media educators to think…”
Of data and transparency
(2023) · Nature Computational Science · doi:10.1038/s43588-023-00499-y
Strengthening the Image of Islamic Schools Through the AIDA Approach: The Role of Public Relations in Engaging the Community
Tantowi, A., Subchan, A., Munadirin, A., Kadarismanto, K., & Muslim, R. (2023) · AL-TANZIM JURNAL MANAJEMEN PENDIDIKAN ISLAM · doi:10.33650/al-tanzim.v7i4.5507
Analyzing the Effectiveness of Image Augmentation for Soybean Crop and Broadleaf Weed Classification
Michael, J. J., & Thenmozhi, M. (2024) · Lecture notes in networks and systems · doi:10.1007/978-981-99-9521-9_27

Authors: Crawford, K., & Paglen, T.
Affiliations: University of Southern California (US)
University of Georgia (US)
Corresponding author: Crawford, Kate
ORCID: Crawford, Kate
Venue: AI & SOCIETY
Publisher: Springer Science and Business Media LLC
Published: 2021-11-23
Volume: 36
Issue: 4
Pages: 1399-1399
ISSN: 0951-5666 (ISSN-L)
1435-5655
DOI: 10.1007/s00146-021-01301-1
OpenAlex: W3217570815
Cited by: 153
References: 3
Citations by year: 2026: 48 · 2025: 49 · 2024: 36 · 2023: 17 · 2022: 3
Type: journal article
Language: en
Open access: bronze
doi.org
Licence: www.springer.com/tdm (tdm)
www.springer.com/tdm (vor)
Locations: AI & Society (publishedVersion)
Topics: Social Sciences › Safety Research › Ethics and Social Impacts of AI
Physical Sciences › Computer Science › Artificial Intelligence › Law, AI, and Intellectual Property
Social Sciences › Law › Digitalization, Law, and Regulation
Keywords: artificial intelligence, machine learning training data, image politics, data curation

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

Correction to: Excavating AI: the politics of images in machine learning training sets

Integrity signals

Claims the paper makes

Self-check

Cites

Cited by

Related

Record

About these checks