Study Proposes Integrative AI Governance Model for Health Systems
A systematic review by Hassane Alami and colleagues, posted as a JMIR preprint (#87448), synthesizes AI governance frameworks for health systems published between November 2014 and July 2025 and proposes an Integrative AI Governance Model to guide policy, practice, and research. The authors searched eight academic databases, grey literature, and international-organization portals, and report that many existing frameworks fail to capture the multidimensional, dynamic nature of AI governance in health systems. Their proposed model organizes governance into structures, processes, and relational mechanisms spanning international, national, local, and organizational levels. A complementary Duke-Margolis white paper frames health-system AI governance as balancing innovation, accountability, and trust, drawing on a multi-stakeholder working group. Together the documents map recurring governance components into an integrative structure. The work is a conceptual synthesis under review, not an empirical evaluation of deployed AI systems.
What happened
Per a JMIR preprint by Hassane Alami and colleagues (#87448), the authors conducted a systematic review of AI governance frameworks for health systems covering November 2014 to July 2025 and propose an Integrative AI Governance Model intended to guide AI-related policy, practice, and research. The preprint reports that existing frameworks aim to address risks such as bias, compromised quality of care, data breaches, and adverse financial impacts, but frequently fail to reflect the multidimensional and dynamic characteristics of AI governance in health systems.
Methods and scope
Per the preprint's methods, the review drew on eight academic databases (including PubMed, MEDLINE, Embase, ACM Digital Library, Web of Science, Scopus, Social Sciences Abstracts, and PsycINFO), grey-literature sources, and international-organization portals. The proposed model organizes governance into structures, processes, and relational mechanisms distributed across international, national, local, and organizational levels. The manuscript synthesizes conceptual elements across frameworks rather than presenting a new empirical evaluation of deployed AI systems.
Industry context
The Duke-Margolis white paper, "AI Governance in Health Systems," frames governance as aligning innovation, accountability, and trust, and documents contributions from a multi-stakeholder working group of health-system privacy, data-science, and engineering leaders. Its emphasis on operational levers - cross-functional working groups, risk-assessment workflows, and alignment with existing clinical governance - maps onto many components the preprint synthesizes.
Editorial analysis
Synthesizing fragmented guidance into an actionable model is a common step once high-level principles proliferate. Health systems operationalizing AI governance typically confront integration problems across data governance, clinical validation, procurement, and regulatory compliance; a single consolidated framework can aid comparative evaluation and policy design.
What to watch
Track whether the preprint completes peer review in the Journal of Medical Internet Research and whether case studies or toolkits operationalize the model. Practical signals of uptake include published implementation studies, institutional policies aligned to the model's domains, and artifacts such as checklists, templates, or evaluation rubrics.
Limitations noted in sources
The work is presented as a preprint under review and synthesizes existing frameworks rather than reporting new empirical validation. The Duke-Margolis white paper is a practitioner-oriented synthesis with invited working-group contributors and was funded by Duke AI Health, as stated in that document.
Key Points
- 1A systematic review (Nov 2014-July 2025, eight databases) consolidates fragmented AI governance frameworks into one comparable structure for health systems.
- 2The proposed model spans structures, processes, and relational mechanisms across international, national, local, and organizational levels.
- 3It is a preprint conceptual synthesis, not empirical validation; uptake depends on peer review and practical toolkits such as checklists or rubrics.
Scoring Rationale
This is a not-yet-peer-reviewed systematic review and conceptual model, useful as a consolidated reference for health-system AI governance teams but not an empirical study or regulatory action. It sits in the interesting-research band: relevant to governance and policy practitioners, with real-world impact contingent on peer review and adoption. Scored below the original 7.1 to reflect its preprint, synthesis-only nature.
Sources
Public references used for this report.
Practice with real Health & Insurance data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Health & Insurance problems