Responsible AI guidance outline
Starter scaffold for the core guidance document that will later consolidate reviewable recommendations.
Purpose
This page is the initial scaffold for a reusable guidance document on responsible generative AI use in biomedical research settings.
Planned contents
- what kinds of tasks are low-risk, cautionary or unsuitable
- how to think about sensitive or human-derived data
- when documentation of AI use should be expected
- where human review remains mandatory
- how local governance and IT constraints override convenience
Current gap
The structure exists, but the guidance should not be finalised until WP1 evidence and literature review findings are integrated.