Making Generative AI Use Visible in a Foundational Health-Sciences Course: Effects on Learning, Verification and Attribution.

Authors

Xinhui Li

Published in

Medical science educator. Volume 36. Issue 5. Pages 2575-2584. Epub Jul 31, 2026.

Abstract

Generative artificial intelligence (GenAI) is increasingly used in health-sciences education, but its benefits in assessed coursework depend on whether students use it in accountable and educationally productive ways. In foundational courses, educators need practical alternatives to both unrestricted use and blanket prohibition. We implemented a course-based GenAI support package in an anatomy and physiology course that aimed to make students' AI use visible through four components: risk-tier guidance, a GenAI use statement, a verification log, and instructor checkpoints. We evaluated the package using a quasi-experimental mixed-methods design with two undergraduate biomedical engineering classes (intervention n = 34; control n = 33). Outcomes included three rubric-scored course reports, a final ethics-focused assessment, two objective structured practical examination stations, survey-based responsible-use indicators, and semi-structured interviews. Compared with business-as-usual teaching, students in the intervention class had higher cumulative report performance (21.59 [3.56] vs. 17.91 [4.12] out of 30; Cohen's d = 0.96) and stronger performance on the final ethics-focused assessment (14.53 [2.11] vs. 12.36 [2.33] out of 20; d = 0.98). They also showed stronger disclosure and attribution practices and improved verification behaviours across repeated coursework. Qualitative findings suggested that making AI use visible helped students clarify boundaries of acceptable use, check claims more systematically, and connect GenAI-supported drafting to course-relevant reasoning and evidence use. A course-based support package that makes GenAI use visible may offer health-sciences educators a feasible way to integrate GenAI into assessed coursework while strengthening verification, attribution and learning, rather than relying on blanket bans alone.
The online version contains supplementary material available at https://doi.org/10.1007/s40670-026-02848-8.

PMID:
42968821
Bibliographic data and abstract were imported from PubMed on 02 Nov 2026.

Read full publication at:
Please sign in to see all details.

Stats

1-terrible, 9-excellent. How would you rate this publication? Sign in in to submit your rating.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.