The DOI matches its Crossref record; 11 of the 11 references it cites were found in a registry or an index; 2 of the 9 sources its claims cite were checked for support.
ChatGPT: five priorities for research
Open accessAbstract
The authors argue that conversational AI will reshape research and outline priorities around verification, accountability, openness, and equity.
Overview
CoChat Summary
TL;DR
This Nature Comment argues that ChatGPT and other large language models will significantly reshape research, but only if the community addresses risks around accuracy, transparency, accountability, and inequality. The authors propose five priorities for research: human verification, accountability rules, open LLMs, embracing benefits, and widening the debate.
Key findings
- The authors report that ChatGPT produced false or misleading outputs on research tasks; in one example, it gave an overly general answer about depression relapse, despite studies showing relapse risk ranges from 29% to 51% in the first year after treatment completion.
- When asked to summarize a systematic review on CBT for anxiety-related disorders, ChatGPT said the review was based on 46 studies; the paper states it was actually based on 69 and says ChatGPT also exaggerated CBT’s effectiveness.
- The article argues that LLMs can speed up tasks such as writing, literature review, coding, and potentially peer review, but can also spread misinformation, bias, and plagiarism.
- The authors recommend five priorities: hold on to human verification, develop rules for accountability, invest in truly open LLMs, embrace the benefits of AI, and widen the debate.
Methods
- This is a Comment/opinion piece, not an empirical study.
- The authors use illustrative examples and argumentation, including direct prompts to ChatGPT and comparison with published research findings.
- No formal sample, experiment design, or statistical analysis is stated in the paper.
Limitations
- The paper does not present a systematic evaluation of ChatGPT; its evidence is based on selected examples and illustrative cases. *(evident from the text)*
- It is a Comment rather than original research, so its claims are interpretive and policy-oriented rather than experimentally tested. *(evident from the text)*
- The paper explicitly notes that its list of discussion questions is non-exhaustive, indicating the framework is incomplete.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
2 Valid · 0 Not valid · 5 Not checked · 13 Own results and interpretations
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Repeating the same query generated a more detailed and accurate answer (see Supplementary information, FigsS1 andS2).”
plainstates a causeSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For example, it said the review was based on 46studies (it was actually based on 69)”
plainSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“and, more worryingly, it exaggerated the effectiveness of CBT.”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For example, we prompted ChatGPT to explain the work of several researchers. In some instances, it produced detailed accounts of scientists who could be considered less influential on the basis of their h-index (a way of measuring the impact of their work).”
hedgedstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“We presented ChatGPT with a series of questions and assignments that required an in-depth understanding of the literature and found that it often generated false and misleading text.”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
3 more claims
Create a free account to see all 20 claims and to verify the 2 references still open.
- It may state two separate claims as one. 8
- Probably not a substantial claim. 4
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?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Its cited benefits included enhancing personalized learning experiences, potentially enhancing communication skills, and increasing students’ engagement in the learning process [9,18,38,39].”
From Assessing Health Students' Attitudes and Usage of ChatGPT in Jordan: Validation Study (2023) · cites this paper as [38]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“This raises ethical issues and goes against the move toward transparency and open science [39].”
From Challenges and Opportunities of Generative AI for Higher Education as Explained by ChatGPT (2023) · cites this paper as [39]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“HAKI's approach aligns with emerging guidelines from leading academic publishers, which emphasize AI disclosure and authorship clarity [86].”
From Developing the HAKI model as a conceptual framework for AI-assisted academic research (2026) · cites this paper as [86]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Moreover, a recent study reported that students believe that the use of ChatGPT can help them receive an education tailored to their needs with immediate feedback, especially in research [31].”
From Health profession students' perceptions of ChatGPT in healthcare and education: insights from a mixed-methods study (2025) · cites this paper as [31]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“This recursive, sometimes multi-round, verification process guarantees that nothing from the AI slips into the final work without human confirmation [86].”
From Developing the HAKI model as a conceptual framework for AI-assisted academic research (2026) · cites this paper as [86]
Experimental — accuracy not yet measured
16 more claims about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
Rudin, C. (2019) · Nature Machine Intelligence · doi:10.1038/s42256-019-0048-x · Cited by 10,561
A constant error in psychological ratings.
Thorndike, E. L. (1920) · Journal of Applied Psychology · doi:10.1037/h0071663 · Cited by 2,163
Partners in Cognition: Extending Human Intelligence with Intelligent Technologies
SALOMON, G., PERKINS, D. N., & GLOBERSON, T. (1991) · Educational Researcher · doi:10.3102/0013189x020003002 · Cited by 948
Vittengl, J. R., Clark, L. A., Dunn, T. W., & Jarrett, R. B. (2007) · Journal of Consulting and Clinical Psychology · doi:10.1037/0022-006x.75.3.475 · Cited by 551
van Dis, E. A. M., van Veen, S. C., Hagenaars, M. A., Batelaan, N. M., Bockting, C. L. H., van den Heuvel, R. M., Cuijpers, P., & Engelhard, I. M. (2020) · JAMA Psychiatry · doi:10.1001/jamapsychiatry.2019.3986 · Cited by 335
Self-citationAccountability and automation bias
SKITKA, L. J., MOSIER, K., & BURDICK, M. D. (2000) · International Journal of Human-Computer Studies · doi:10.1006/ijhc.1999.0349 · Cited by 324
Active learning machine learns to create new quantum experiments
Melnikov, A. A., Poulsen Nautrup, H., Krenn, M., Dunjko, V., Tiersch, M., Zeilinger, A., & Briegel, H. J. (2018) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1714936115 · Cited by 296
Relapse Following Successful Electroconvulsive Therapy for Major Depression: A Meta-Analysis
Jelovac, A., Kolshus, E., & McLoughlin, D. M. (2013) · Neuropsychopharmacology · doi:10.1038/npp.2013.149 · Cited by 287
Kato, M., Hori, H., Inoue, T., Iga, J., Iwata, M., Inagaki, T., Shinohara, K., Imai, H., Murata, A., Mishima, K., et al. (2021) · Molecular Psychiatry · doi:10.1038/s41380-020-0843-0 · Cited by 152
Lessons for artificial intelligence from the study of natural stupidity
Rich, A. S., & Gureckis, T. M. (2019) · Nature Machine Intelligence · doi:10.1038/s42256-019-0038-z · Cited by 94
Cited by
Dwivedi, Y. K., Kshetri, N., Hughes, L., Slade, E. L., Jeyaraj, A., Kar, A. K., Baabdullah, A. M., Koohang, A., Raghavan, V., Ahuja, M., et al. (2023) · International Journal of Information Management · doi:10.1016/j.ijinfomgt.2023.102642 · Cited by 4,584
Sallam, M. (2023) · Healthcare · doi:10.3390/healthcare11060887 · Cited by 3,126
Ray, P. P. (2023) · Internet of Things and Cyber-Physical Systems · doi:10.1016/j.iotcps.2023.04.003 · Cited by 2,463
Students’ voices on generative AI: perceptions, benefits, and challenges in higher education
Chan, C. K. Y., & Hu, W. (2023) · International Journal of Educational Technology in Higher Education · doi:10.1186/s41239-023-00411-8 · Cited by 2,301
“…to be useful research aids for generating ideas, synthesizing information, and summarising a vast amount of text data to help researchers analyse data and compose their writing (Berg, 2023; Chan & Zhou, 2023), contributing to efficiency in publication (Kitamura, 2023; van Dis et al., 2023).”
BackgroundGenerative AI and ChatGPT: applications, challenges, and AI-human collaboration
Fui-Hoon Nah, F., Zheng, R., Cai, J., Siau, K., & Chen, L. (2023) · Journal of Information Technology Case and Application Research · doi:10.1080/15228053.2023.2233814 · Cited by 1,363
Metadata
- Authors
- van Dis, E. A. M., Bollen, J., Zuidema, W., van Rooij, R., & Bockting, C. L.
- Affiliations
- Indiana University Bloomington (US)
- University of Amsterdam (NL)
- Amsterdam University Medical Centers (NL)
- Corresponding author
- Bockting, Claudi L.
- Venue
- Nature
- Published
- 2023-02-03
- Volume
- 614
- Issue
- 7947
- Pages
- 224-226
- ISSN
- 0028-0836 (ISSN-L)
- 1476-4687
- OpenAlex
- W4319083882
- Cited by
- 1,842
- References
- 11
- Citations by year
- 2026: 253 · 2025: 407 · 2024: 589 · 2023: 588
- FWCI
- 172.70
- Type
- journal article
- Language
- en
- Open access
- green
- pure.amsterdamumc.nl
- Licence
- cc-by
- www.springernature.com/gp/researchers/text-and-data-mining (tdm)
- www.springernature.com/gp/researchers/text-and-data-mining (vor)
- Locations
- Nature (publishedVersion)
- PubMed (publishedVersion)
- Pure Amsterdam UMC (submittedVersion, cc-by)
- RePEc: Research Papers in Economics (submittedVersion)
- UvA-DARE (University of Amsterdam) (submittedVersion)
- Data Archiving and Networked Services (DANS) (submittedVersion)
- Funders
- Nederlandse Organisatie voor Wetenschappelijk Onderzoek (406.18.TW.007)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Artificial Intelligence Applications
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Keywords
- ChatGPT, conversational AI, large language models, research priorities, artificial intelligence in science, research ethics
- MeSH terms
- Machine Learning, Research, Writing
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.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
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: ChatGPT: five priorities for research. https://app.cochat.ai/papers/10.1038/d41586-023-00288-7
- The paper
- van Dis, E. A. M., Bollen, J., Zuidema, W., van Rooij, R., & Bockting, C. L. (2023). ChatGPT: five priorities for research. Nature. https://doi.org/10.1038/d41586-023-00288-7
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.