Work

Writing with the Machine: Cautious Adoption, Accuracy Concerns, and the Policy Vacuum in Health Sciences Grant Writing

Year: 2026
Type: preprint
Study design: Observational Study
Abstract: This study examined how health sciences faculty use, need, and evaluate generative artificial (GenAI) intelligence tools in grant writing. A mixed-methods study, it used an anonymous cross-sectional survey administered to faculty across seven health sciences colleges at a large public research university in the United States. Ninety responses were retained for analysis from 2,119 invited faculty (4.2% response rate). Quantitative data from the survey were analyzed descriptively, and open-ended responses were examined using inductive qualitative content analysis. More than half of respondents (54.4%) reported using GenAI tools in grant writing, primarily for editing, rewording, organization, and language refinement. Accuracy concerns were widespread, with 75.3% reporting concern always or most of the time; 70.0% reported ethical or authorship concerns, and 90.2% identified accuracy or reliability as a barrier to use. Institutional guidance was perceived as limited, with 52.8% disagreeing that clear guidance existed. Qualitative findings identified five themes: language-level assistance, verification burden, intellectual ownership and voice, ethical and policy uncertainty, and moral or environmental objections to AI use. Findings suggest cautious, selective adoption of GenAI and identify opportunities for libraries to support AI literacy, citation verification, policy interpretation, research integrity, and responsible AI use in grant development.
Language: English
Subfield: Education
Open Access status: green
OpenAlex ID: W7226014628
Cites:
Cited by:
Related to: 10