Responsible use of AI in evidence SynthEsis (RAISE): recommendations and guidance
Open accessAbstract
** If you would like to provide feedback on RAISE, please use this feedback form https://forms.office.com/e/EDYvBPEBrd ** The RAISE guidance is currently written in three documents (click 'Files' in the menu above). The first document, ‘RAISE 1’, provides tailored recommendations for eight of the distinct roles in the evidence synthesis ecosystem: evidence synthesists, methodologists, AI tool developments teams, organisations that produce evidence synthesis, publishers, funders, users, and trainers of evidence synthesis methods. RAISE 2 contains guidance on building and evaluating AI evidence synthesis tools, which focuses on determining if an AI tool does what it claims to do to an acceptable standard, including how to build and validate AI tools, conduct evaluations to build a cumulative evidence base, including performance metrics to consider, and report evaluations. The final document, RAISE 3, then offers guidance on selecting and using AI evidence synthesis tools, which focuses on understanding whether an AI tool can be used for a specific evidence synthesis, including how to assess, select and use an AI tool, including ethical, legal and regulatory considerations, and the current state of AI tools for evidence synthesis. Please see the storage area for the RAISE documents. https://osf.io/fwaud/files/osfstorage We invite feedback and thoughts on this work. Please use https://forms.office.com/e/EDYvBPEBrd to submit feedback on RAISE.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
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?
Human oversight is essential to ensure trust in both the evidence synthesis and the conclusions made.
“This is essential to ensure trust in both the evidence synthesis and the conclusions made [111], and to guard against the potential responsibility gap if errors do occur.”
From Comparison of Elicit AI and Traditional Literature Searching in Evidence Syntheses Using Four Case Studies (2025) · cites this paper as [111]
Cites
Still being checked.
Cited by
Sorted by citations
Opportunities, challenges and risks of using artificial intelligence for evidence synthesis
Siemens, W., von Elm, E. B., Binder, H., Böhringer, D., Eisele‐Metzger, A., Gartlehner, G., Hanegraaf, P., Metzendorf, M., Mosselman, J. J., Nowak, A. J., et al. (2025) · BMJ Evidence-Based Medicine · doi:10.1136/bmjebm-2024-113320 · Cited by 34 · see citing works
Self-citationFlemyng, E., Noel-Storr, A. H., Macura, B., Gartlehner, G., Thomas, J., Meerpohl, J. J., Jordan, Z., Minx, J. C., Eisele‐Metzger, A., Hamel, C., et al. (2025) · Environmental Evidence · doi:10.1186/s13750-025-00374-5 · Cited by 21 · see citing works
Self-citationLau, O., & Golder, S. (2025) · Cochrane Evidence Synthesis and Methods · doi:10.1002/cesm.70050 · Cited by 11 · see citing works
Gartlehner, G., Nußbaumer-Streit, B., Hamel, C., Garritty, C., Griebler, U., King, V. J., Devane, D., & Kamel, C. (2025) · Cochrane Evidence Synthesis and Methods · doi:10.1002/cesm.70063 · Cited by 12 · see citing works
Lagisz, M., Mizuno, A., Morrison, K., Pollo, P., Ricolfi, L., Yang, Y., & Nakagawa, S. (2026) · Research Synthesis Methods · doi:10.1017/rsm.2026.10080 · Cited by 10 · see citing works
Fontaine, G., Michie, S., Beidas, R. S., Geng, E. H., Fahim, C., Powell, B. J., Welch, V. N., Thomas, J., Chan, J., Abbasgholizadeh-Rahimi, S., et al. (2026) · Implementation Science · doi:10.1186/s13012-026-01503-5 · Cited by 5 · see citing works
Self-citationHilkenmeier, F., Stoltenberg, M., & Stierle, C. M. G. (2026) · Systematic Reviews · doi:10.1186/s13643-026-03228-4 · Cited by 5 · see citing works
Jansen, T., Liebenow, L. W., Schaller, N.-J., Hattie, J., & Möller, J. (2026) · Educational Psychology Review · doi:10.1007/s10648-026-10136-5 · Cited by 1 · see citing works
‘To Boldly Go…’ Guidance for Engineering Librarians Navigating the Landscape of Knowledge Synthesis
Mawhinney, T., & Morris, M. (2026) · Science & Technology Libraries · doi:10.1080/0194262x.2026.2674695
Response to: Five methodological considerations for validating LLMs in risk of bias assessment
Eisele‐Metzger, A., Lieberum, J.-L., Toews, M., Siemens, W., Heilmeyer, F., Haverkamp, C., Böhringer, D., & Meerpohl, J. J. (2026) · Research Synthesis Methods · doi:10.1017/rsm.2026.10102
Jóźwiak, Á. B., Imre, A., Józwiak‐Hagymásy, J., Tittmann, J., Nagy, Á., Kovács, S., Kardas, P., van Boven, J. F. M., Mommers, I., & Ágh, T. (2026) · Expert Review of Pharmacoeconomics & Outcomes Research · doi:10.1080/14737167.2026.2691995
Librarians and information specialists: Partners in evidence-based toxicology and other disciplines
Price, C. L., & Rethlefsen, M. L. (2026) · Evidence-Based Toxicology · doi:10.1080/2833373x.2026.2728359
Yin, C., Jing, Z., & Zhang, Z. (2026) · medRxiv · doi:10.64898/2026.09.21.26363538
Weiss, B., Wagner, M., Daikeler, J., & Matthias, K. (2026) · Springer eBooks · doi:10.1007/978-3-658-52312-1_59
Kattimani, V. S. (2026) · Journal of Maxillofacial and Oral Surgery · doi:10.1007/s12663-026-03312-1
- Authors
- Thomas, J., Flemyng, E., & Noel-Storr, A.
- Venue
- OSF
- Publisher
- OSF
- Published
- 2025
- OpenAlex
- W6943748579
- Cited by
- 14
- References
- 0
- Citations by year
- 2026: 11 · 2025: 3
- Type
- posted content
- Language
- en
- Open access
- green
- osf.io
- Licence
- Locations
- OSF Preprints (OSF Preprints) (submittedVersion, cc-by)
- Open Science Framework (cc-by)
- Topics
- Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
- Social Sciences › Safety Research › Ethics and Social Impacts of AI
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Keywords
- responsible AI use, AI tool evaluation, performance indicators, AI ethics
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
