The DOI matches its Crossref record; 10 of the 10 references it cites were found in a registry or an index; none of the 3 sources its claims cite is checked for support yet.
Protecting scientific integrity in an age of generative AI
Open accessAbstract
Competing interest statement: V.G.C., M.L
Overview
CoChat Summary
TL;DR
This PNAS article calls on the scientific community to protect scientific integrity amid generative AI by upholding human accountability and responsibility. It proposes five principles for AI use in science and recommends that the National Academies establish a Strategic Council on the Responsible Use of Artificial Intelligence in Science.
Key findings
- Generative AI challenges core scientific norms—accountability, transparency, replicability, and human responsibility—and makes it harder to confirm veracity, attribute machine- versus human-authored content, disclose AI use, replicate studies, and identify biases and inequities.
- The authors urge adoption of five principles: transparent disclosure and attribution; verification of AI-generated content and analyses; documentation of AI-generated data; a focus on ethics and equity; and continuous monitoring, oversight, and public engagement.
- They propose that the National Academies of Sciences, Engineering, and Medicine establish a Strategic Council on the Responsible Use of Artificial Intelligence in Science to coordinate guidance, monitor evolving uses, address ethical and societal concerns, and refine best practices.
- The article reaffirms prior National Academies work on responsible computing research and automated research workflows, including calls for human review of algorithms, transparency, reproducibility, and efforts to uncover and address bias.
Methods
- This is a position/commentary piece, not an empirical study. It reports on an interdisciplinary panel convened by the National Academy of Sciences, the Annenberg Public Policy Center of the University of Pennsylvania, and the Annenberg Foundation Trust at Sunnylands.
- The panel included experts in behavioral and social sciences, ethics, biology, physics, chemistry, mathematics, computer science, higher education, law, governance, and science publishing and communication.
- Discussions were informed by commissioned papers on AI technologies, effects on equality/justice/research ethics, governance issues, and historical lessons; the article synthesizes these into principles and a governance proposal.
Limitations
- The paper states no formal limitations.
- Evident from the text: the recommendations are consensus-based and normative, with no empirical data, systematic review, implementation metrics, or evaluation of effectiveness.
- The authors note AI technologies are rapidly evolving, which implies guidance will need continuous updating; they propose the Strategic Council partly for that reason.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
0 Valid · 0 Not valid · 1 Not checked · 7 Own results and interpretations
A claim is checked only against the cited paper’s full text, never its abstract, and none is available for its sources.
In the paper
“its tools and processes challenge core norms and values in the conduct of science, including accountability, transparency, replicability, and human responsibility (1–3).”
What the sources say
- Not checked — no full text of the sourcecites [1] National Academies of Sciences, Engineering, and Medicine. Fostering Integrity in Research (The National Academies Press, Washington, DC, 2017), 10.17226/21896. 29341557 source
- Not checked — no full text of the sourcecites [2] National Academies of Sciences, Engineering, and Medicine. Reproducibility and Replicability in Science (The National Academies Press, Washington, DC, 2019), 10.17226/25303. 31596559 source
- Not checked — no full text of the sourcecites [3] National Academies of Sciences, Engineering, and Medicine. Fostering Responsible Computing Research: Foundations and Practices (The National Academies Press, Washington, DC, 2022), 10.17226/26507. source
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: unclear — It is unclear whether it could be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Probably not a substantial claim. 7
- The quote only partly states this claim. 1
Citations
Cites
Reproducibility and Replicability in Science
Committee on Reproducibility and Replicability in Science, Board on Behavioral, C., Committee on National Statistics, Education, D. O. B. A. S. S. A., Nuclear and Radiation Studies Board, Studies, D. O. E. A. L., Board on Mathematical Sciences and Analytics, Committee on Applied and Theoretical Statistics, Sciences, D. O. E. A. P., Board on Research Data and Information, et al. (2019) · doi:10.17226/25303 · Cited by 924
Fostering Integrity in Research
Committee on Responsible Science, Committee on Science, E., Affairs, P. A. G., & National Academies of Sciences, E. (2017) · doi:10.17226/21896 · Cited by 186
FOSTERING INTEGRITY IN RESEARCH
Arrison, T., & Nerem, R. M. (2017) · Special publications · doi:10.1002/9781119067825.ch2 · Cited by 136
Automated Research Workflows for Accelerated Discovery
Board on Mathematical Sciences and Analytics, Computer Science and Telecommunications Board, Sciences, D. O. E. A. P., et al. (2022) · National Academies Press eBooks · doi:10.17226/26532 · Cited by 18
Fostering Responsible Computing Research: Foundations and Practices
Committee on Responsible Computing Research: Ethics and Governance of Computing Research and Its Applications, Computer Science and Telecommunications Board, Sciences, D. O. E. A. P., & National Academies of Sciences, E. (2022) · doi:10.17226/26507 · Cited by 10
Bringing Communities In, Achieving AI for All
Parthasarathy, S., & Katzman, J. (2024) · Issues in Science and Technology · doi:10.58875/slrg2529 · Cited by 7
Self-citationA Human Rights Framework for AI Research Worthy of Public Trust
Gray, M. W. (2024) · Issues in Science and Technology · doi:10.58875/eruu8159 · Cited by 3
Self-citationGoverning AI With Intelligence
Gasser, U. (2024) · Issues in Science and Technology · doi:10.58875/awjg1236 · Cited by 2
Self-citationAidinoff, M., & Kaiser, D. I. (2024) · Issues in Science and Technology · doi:10.58875/buxb2813 · Cited by 1
Self-citationA Justice-Led Approach to AI Innovation
London, A. J. (2024) · Issues in Science and Technology · doi:10.58875/knrz2697 · Cited by 1
Self-citation
Research works printed but not linked: 1
- E. Horvitz T. M. Mitchell “Progress in AI: History and Trajectory” in Issues in Science and Technology (in press) 10.58875/ZRFJ6746. · doi:10.58875/zrfj6746
Cited by
Obschonka, M., Grégoire, D. A., Nikolaev, B., Ooms, F., Lévesque, M., Pollack, J. M., & Behrend, T. S. (2024) · Entrepreneurship Theory and Practice · doi:10.1177/10422587241304676 · Cited by 77
The role of generative AI in academic and scientific authorship: an autopoietic perspective
Watson, S., Brezovec, E., & Romic, J. (2025) · AI & Society · doi:10.1007/s00146-024-02174-w · Cited by 78
“This is often formulated in terms of a defense of integrity, authenticity and the privileging of human rationality (Blau et al 2024; Bagenal 2024).”
Exploring the scope of generative AI in literature review development
Schryen, G., Marrone, M., & Yang, J. (2025) · Electronic Markets · doi:10.1007/s12525-025-00754-2 · Cited by 71
“As GenAI increasingly integrates into LR processes, it is crucial to adhere to the core values of accountability, transparency, replicability, and responsibility (Blau et al., 2024).”
A call for built-in biosecurity safeguards for generative AI tools
Wang, M., Zhang, Z., Bedi, A. S., Velasquez, A., Guerra, S., Lin‐Gibson, S., Cong, L., Qu, Y., Chakraborty, S., Blewett, M. M., et al. (2025) · Nature Biotechnology · doi:10.1038/s41587-025-02650-8 · Cited by 58
Rise of Generative Artificial Intelligence in Science
Ding, L., Lawson, C., & Philip Shapira, A. (2025) · Scientometrics · doi:10.1007/s11192-025-05413-z · Cited by 45
“However, notwithstanding its rapid spread, the deployment of generative AI remains fraught with ethical and epistemic challenges (Blau et al., 2024; Charness et al., 2023).”
Metadata
- Authors
- Blau, W., Cerf, V. G., Enriquez, J., Francisco, J. S., Gasser, U., Gray, M. L., Greaves, M., Grosz, B. J., Jamieson, K. H., Haug, G. H., Hennessy, J. L., Horvitz, E., Kaiser, D. I., London, A. J., Lovell-Badge, R., McNutt, M. K., Minow, M., Mitchell, T. M., Ness, S., Parthasarathy, S., Perlmutter, S., Press, W. H., Wing, J. M., & Witherell, M.
- Affiliations
- Brunswick Group, London WC2A 3ED, United Kingdom
- Google (United States) (US)
- Excel Venture Management, Boston, MA 02116
- University of Pennsylvania (US)
- Technical University of Munich (DE)
- Indiana University Bloomington (US)
- Schmidt Sciences (US)
- Harvard University (US)
- German National Academy of Sciences Leopoldina (DE)
- Stanford University (US)
- Microsoft (United States) (US)
- Massachusetts Institute of Technology (US)
- Carnegie Mellon University (US)
- The Francis Crick Institute (GB)
- National Academy of Sciences (US)
- University of Michigan (US)
- Lawrence Berkeley National Laboratory (US)
- The University of Texas at Austin (US)
- Columbia University (US)
- Corresponding author
- McNutt, Marcia K.
- ORCID
- Cerf, Vinton G.
- Francisco, Joseph S.
- Gasser, Urs
- Gray, Mary L.
- Grosz, Barbara J.
- Jamieson, Kathleen Hall
- Haug, Gerald H.
- Hennessy, John L.
- Horvitz, Eric
- Kaiser, David I.
- London, Alex John
- Lovell-Badge, Robin
- McNutt, Marcia K.
- Mitchell, Tom M.
- Parthasarathy, Shobita
- Perlmutter, Saul
- Press, William H.
- Wing, Jeannette M.
- Witherell, Michael
- Venue
- Proceedings of the National Academy of Sciences
- Publisher
- National Academy of Sciences
- Published
- 2024-05-21
- Volume
- 121
- Issue
- 22
- Pages
- e2407886121-e2407886121
- ISSN
- 0027-8424 (ISSN-L)
- 1091-6490
- PMC
- PMC11145223
- OpenAlex
- W4398183095
- Cited by
- 118
- References
- 7
- Citations by year
- 2026: 45 · 2025: 58 · 2024: 15
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- Proceedings of the National Academy of Sciences (publishedVersion, cc-by-nc-nd)
- PubMed Central (publishedVersion, cc-by-nc-nd)
- PubMed (publishedVersion)
- eScholarship (California Digital Library) (submittedVersion, cc-by)
- OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) (submittedVersion)
- Repository for Publications and Research Data (ETH Zurich) (submittedVersion, cc-by-nc-nd)
- ETH Zürich Research Collection
- Funders
- California Institute of Technology
- Weill Cornell Medical College
- Topics
- Social Sciences › Education › Academic integrity and plagiarism
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Social Sciences › Library and Information Sciences › Academic Publishing and Open Access
- Keywords
- scientific integrity, generative AI, artificial intelligence in research, research ethics, AI-generated content, research misconduct
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.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
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: Protecting scientific integrity in an age of generative AI. https://app.cochat.ai/papers/10.1073/pnas.2407886121
- The paper
- Blau, W., Cerf, V. G., Enriquez, J., Francisco, J. S., Gasser, U., Gray, M. L., Greaves, M., Grosz, B. J., Jamieson, K. H., Haug, G. H., Hennessy, J. L., Horvitz, E., Kaiser, D. I., London, A. J., Lovell-Badge, R., McNutt, M. K., Minow, M., Mitchell, T. M., Ness, S., . . . Witherell, M. (2024). Protecting scientific integrity in an age of generative AI. Proceedings of the National Academy of Sciences. https://doi.org/10.1073/pnas.2407886121
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.