An international team of researchers, clinicians, educators, ethicists, and patient partners released the Health CARE-AI Framework on July 27, 2026, a consensus-backed set of 10 principles that defines professionalism expectations for AI across health education, research, and patient care. The framework, published in JMIR Medical Education, was validated through a rigorous 3-phase modified Delphi process with 303 unique participants. It arrives as health systems worldwide struggle to translate broad ethical AI guidelines into day-to-day clinical and educational practice.
While organizations like the WHO and UNESCO have established high-level principles, frontline clinicians and educators often lack clear direction on concrete issues-patient privacy, algorithmic bias, student assessment, and data stewardship. The framework addresses that gap by moving beyond abstract values into competencies that can be taught, supervised, and institutionalized. It is designed specifically for those integrating AI for Healthcare, where the stakes involve clinical autonomy, relational trust, and health equity.
From high-level principles to daily accountability
The Health CARE-AI initiative recognizes that AI is not simply an extension of existing digital tools. It introduces fundamental challenges to how care is delivered and learned. The framework's 10 core principles are organized into four interconnected domains, each targeting a specific layer of responsibility.
The four domains and ten principles
Values establishes AI use as a shared individual and collective duty, demanding transparency, honesty, and integrity in AI-assisted care and learning.
Competence commits to continuous, role-appropriate AI literacy and insists that critical human judgment remains central-AI must complement, not replace, clinical and educational decision-making.
Accountability treats AI as a present "third party" during interactions, enforces strict legal, privacy, and consent boundaries, and requires ethical data stewardship.
Structural Equity focuses on identifying and mitigating algorithmic bias, embedding equity into AI design and governance through co-design with affected communities, and advancing environmental and workforce sustainability.
Designed for immediate real-world use
Unlike static policy documents, the Health CARE-AI Framework is paired with a companion implementation guide and toolkit. The toolkit includes scenario-based applications across teaching, research, and governance. Medical schools, residency programs, and health system leaders can use it to evaluate readiness, update curricula, and audit AI deployments. The consensus was near-unanimous: 96% of Delphi participants agreed or strongly agreed that the framework clearly defines professionalism expectations for AI across educational, technological, and ethical needs.
Why this matters for science and research professionals
For researchers and scientists working with AI in health contexts, the framework provides a validated, operational structure that moves ethics from aspiration to audit. The emphasis on co-design with affected communities and algorithmic bias mitigation aligns directly with best practices in AI for Science & Research. The accompanying toolkit offers a concrete method to evaluate AI tools, design equitable studies, and ensure responsible data stewardship-competencies that are increasingly expected by funders, journals, and institutional review boards.
Your membership also unlocks:
