The DOI matches its Crossref record, with a small difference in title or year; 9 of the 9 references it cites were found in a registry or an index.
Observe, inspect, modify: Three conditions for generative AI governance
Abstract
In a world increasingly shaped by generative AI systems like ChatGPT, the absence of benchmarks to examine the efficacy of oversight mechanisms is a problem for research and policy. What are the structural conditions for governing generative AI systems? To answer this question, it is crucial to situate generative AI systems as regulatory objects: material items that can be governed. On this conceptual basis, we introduce three high-level conditions to structure research and policy agendas on generative AI governance: industrial observability, public inspectability, and technical modifiability. Empirically, we explicate those conditions with a focus on the EU’s AI Act, grounding the analysis of oversight mechanisms for generative AI systems in their granular material properties as observable, inspectable, and modifiable objects. Those three conditions represent an action plan to help us perceive generative AI systems as negotiable objects, rather than seeing them as mysterious forces that pose existential risks for humanity.
Overview
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
Verification
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?
In the citing paper
“These tools can significantly strengthen the effectiveness of viral disinformation campaigns, making it easier to produce and disseminate highly tailored and convincing false information (Pérez Dasilva et al., 2021), targeting specific groups and sparking societal and political tensions (Ferrari et al., 2023).”
From Exploring the Challenges of Generative AI on Public Sector Communication in Europe (2025) · cites this paper as Ferrari et al., 2023
In the citing paper
“However, the emergence of AI tools such as ChatGPT has prompted a revision to include such technologies (Ferrari et al., 2023).”
From Exploring the Challenges of Generative AI on Public Sector Communication in Europe (2025) · cites this paper as Ferrari et al., 2023
In the citing paper
“Some interviewees, moreover, do not currently perceive greater ethical risk for communication management, particularly due to the role played by the European Commission in AI regulation (Ferrari et al., 2023), and a greater sense of responsibility that these professionals experience while performing their strategic roles.”
From Exploring the Challenges of Generative AI on Public Sector Communication in Europe (2025) · cites this paper as Ferrari et al., 2023
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
Large Language Models Encode Clinical Knowledge
Singhal, K., Azizi, S., Tu, T., Mahdavi, S. S., Wei, J., Chung, H. W., Scales, N., Tanwani, A., Cole-Lewis, H., Pfohl, S., et al. (2022) · arXiv · doi:10.48550/arxiv.2212.13138 · Cited by 4,083
Large language models encode clinical knowledge
Singhal, K., Azizi, S., Tu, T., Mahdavi, S. S., Wei, J., Chung, H. W., Scales, N., Tanwani, A., Cole-Lewis, H., Pfohl, S., et al. (2023) · Nature · doi:10.1038/s41586-023-06291-2 · Cited by 4,083
How the machine “thinks”: Understanding opacity in machine learning algorithms
Burrell, J. (2016) · Big Data & Society · doi:10.1177/2053951715622512 · Cited by 3,196 · cited 3×
“…oversight approaches, we argue that it is possible to appreciate that generative AI systems are indeed mathematically complex objects that partly transcend human-scale reasoning (Burrell, 2016), while also stressing why their developers may deliberately hide information about those systems.”
(2004) · doi:10.4324/9780203413845 · Cited by 2,963
Model Cards for Model Reporting
Mitchell, M., Wu, S., Zaldivar, A., Barnes, P., Vasserman, L., Hutchinson, B., Spitzer, E., Raji, I. D., & Gebru, T. (2019) · Proceedings of the Conference on Fairness, Accountability, and Transparency · doi:10.1145/3287560.3287596 · Cited by 2,681 · cited 1×
“Model cards, for example, have evolved as bottom-up standardization devices for AI developers to state all essential components of a foundation model, including a specification of (domain-specific) training data sets, but also biases and ethical considerations (Mitchell et al., 2019).”
On 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
Liesenfeld, A., Lopez, A., & Dingemanse, M. (2023) · Proceedings of the 5th International Conference on Conversational User Interfaces · doi:10.1145/3571884.3604316 · Cited by 111
The Gradient of Generative AI Release: Methods and Considerations
Solaiman, I. (2023) · 2023 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3593013.3593981 · Cited by 90
Open (For Business): Big Tech, Concentrated Power, and the Political Economy of Open AI
Widder, D. G., West, S., & Whittaker, M. (2023) · doi:10.2139/ssrn.4543807 · Cited by 87
Research works printed but not linked: 23
- Birhane A Prabhu VU Kahembwe E (2021) Multimodal datasets: misogyny pornography and malignant stereotypes. arXiv. Available at: https://arxiv.org/abs/2110.01963 · arXiv
- 10.1177/2053951720935146 · DOI
- 10.1017/elo.2022.47 · DOI
- 10.1177/26349825231193226 · DOI
- 10.1038/s42256-023-00695-5 · DOI
- 10.1111/reel.12081 · doi:10.1111/reel.12081
- Gillespie T (2018) Custodians of the Internet: Platforms, Content Moderation, and the Hidden Decisions That Shape Social Media. New Haven, CT: Yale University Press.
- Gonen H Goldberg Y (2019) Lipstick on a pig: debiasing methods cover up systematic gender biases in word embeddings but do not remove them. arXiv. Available at: https://arxiv.org/abs/1903.03862 · DOI · Paper page
- 10.1017/9781108890960.013 · DOI
- Helberger N Diakopoulos N (2023) ChatGPT and the AI act. Internet Policy Review 12(1). Available at: https://policyreview.info/essay/chatgpt-and-ai-act · DOI
- 10.1215/9780822383543 · DOI · Paper page
- 10.1177/0306312706059461 · DOI · Paper page
- Li P Yang J Islam MA et al. (2023) Making AI less “thirsty”: uncovering and addressing the secret water footprint of AI models. arXiv preprint. arXiv:2307.05532. Available at: https://arxiv.org/abs/2307.05532 · arXiv
- 10.1177/20539517211047734 · DOI
- 10.1007/s43681-023-00289-2. · DOI · Paper page
- 10.18653/v1/2022.naacl-main.188 · DOI
- 10.14763/2020.4.1535 · DOI · Paper page
- Scao TL Fan A Akiki C et al. (2023) BLOOM: A 176B-Parameter Open-Access Multilingual Language Model. arXiv Preprint. arXiv:2211.05100v4. Available at: https://arxiv.org/abs/2211.05100 · arXiv
- 10.1177/2053951717738104 · DOI · Paper page
- 10.1080/1369118X.2021.1874035 · DOI
- 10.1177/1461444820940293 · DOI · Paper page
- Vaswani A Shazeer N Parmar N et al. (2017) Attention is all you need. Advances in Neural Information Processing Systems. Available at: https://proceedings.neurips.cc/paper_files/paper/2017/file/3f5ee243547dee91fbd053c1c4a845aa-Paper.pdf
- Veale M Matus K Gorwa R (2023) AI and global governance: Modalities rationales tensions. Annual Review of Law and Social Science. SocArXiv preprint. Available at: https://doi.org/10.1146/annurev-lawsocsci-020223-040749 · DOI
Cited by
Ethical challenges and solutions of generative AI: an interdisciplinary perspective
Al-kfairy, M., Mustafa, D., Kshetri, N., Insiew, M., & Alfandi, O. (2024) · Informatics · doi:10.3390/informatics11030058 · Cited by 328
Exploring the Challenges of Generative AI on Public Sector Communication in Europe
Lovari, A., & De Rosa, F. (2025) · Media and Communication · doi:10.17645/mac.9644 · Cited by 22
Kuai, J., Brantner, C., Karlsson, M., Van Couvering, E., & Romano, S. (2026) · New Media & Society · doi:10.1177/14614448251321162 · Cited by 16
“With integrated web-search capabilities, AI chatbots are poised to transform how people access and interact with information, raising urgent questions about the quality of their outputs and the border implications for democratic engage-ments (Ferrari et al., 2023).”
Metadata
- Authors
- Ferrari, F., van Dijck, J., & van den Bosch, A.
- Affiliations
- Utrecht University, The Netherlands
- Venue
- New Media & Society
- Publisher
- SAGE Publications
- Published
- 2023-11-29
- Volume
- 27
- Issue
- 5
- Pages
- 2788-2806
- ISSN
- 1461-4448
- 1461-7315
- Cited by
- 43
- References
- 9
- Language
- en
- Licence
- creativecommons.org/licenses/by/4.0/ (unspecified)
- journals.sagepub.com/page/policies/text-and-data-mining-license (tdm)
- Funders
- Nederlandse Organisatie voor Wetenschappelijk Onderzoek (Spinoza Prize)
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: Observe, inspect, modify: Three conditions for generative AI governance. https://app.cochat.ai/papers/10.1177/14614448231214811
- The paper
- Ferrari, F., van Dijck, J., & van den Bosch, A. (2025). Observe, inspect, modify: Three conditions for generative AI governance. New Media & Society. https://doi.org/10.1177/14614448231214811
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.
