Can Large Language Models support research grant peer review?
Media mentions or quotes:
Quoted in Research Professional article, “Ex-UKRI chief warns against leaning on AI in funding decisions”
Mentioned in Research Professional article, “Funder defends data security after ‘AI-based triage’ row”
Research focus on in the Research Professional article, “AI ‘could filter out weaker grant proposals’, study suggests”.
Science, quoted in, “U.K. funder uses ‘AI triage’ to reject grant proposals—leaving researchers dismayed..
Preprint available:
Thelwall, M., Pride, D., Osborne, F., Salatino, A., Knoth, P., Iyawa, G., & Yusuf, O. (2026). Scoring Grant Applications with Large Language Models. arXiv preprint arXiv:2609.25327.
See also:
Sandström, U. & Thelwall, M. (2026). Can Large Language Models evaluate grant proposal quality? Revisiting the Wennerås and Wold peer review data. Journal of Data and Information Science. https://doi.org/10.1515/jdis-2026-0048
Assess the extent to which different Large Language Scores can predict peer review scores for grant proposals.
Simulate the inclusion of LLM scores in grant peer review pipelines.
Investigate the equality, diversity and inclusion implications of including LLM scores in grant peer review.
Develop a dashboard to support research managers using or evaluating LLM scores to support grant peer review.
This is an ESRC funded metascience project set to run for a year from July 2025. Contact Mike or any other member of the team for more information. Formal start date: 1 October 2025.
Award name:
UKRI2101: LLMs Supporting Grant Peer Review
Opportunity:
OPP934: UKRI Data for Metascience
Application:
UKRI1486: LLMs Supporting Grant Peer Review
We are paired with the AI Peer project.
If you have any concerns about this project, including data processing, please email Mike Thelwall.