Policy Brief No. 217 — November 2025
A Community-Centred Protocol for Ethical and Scalable AI in Health Care Abbas Yazdinejad, Maral Niazi, James W. Hinton, Jude Kong, Jake Okechukwu Effoduh and Anna Shin
Key Points → Punitive intellectual property (IP) frameworks and inadequate data sovereignty protections are significant barriers to equitable artificial intelligence (AI) in health care. These barriers disproportionately affect marginalized populations, necessitating urgent reform. → The authors propose a novel, community-centred AI protocol that integrates FAIR (Findable, Accessible, Interoperable and Reusable) and FHIR (Fast Healthcare Interoperability Resources) standards with flexible IP governance and robust community engagement to address these challenges. → This approach may be applied through a system that manages IP rights to drive public benefit, as well as a data collective that provides managed access to data so as to prevent misappropriation, and thus effective access to important data that would contribute to greater innovation and access to information within health care.
Introduction The integration of AI into health care holds immense promise for enhancing diagnostic accuracy, personalizing treatments as well as optimizing public health systems (Chustecki 2024). However, significant governance challenges threaten to undermine these benefits, particularly for underserved communities. Punitive IP frameworks — overly rigid rules and enforcement practices that prioritize proprietary control over public health needs — restrict access to innovations, especially during crises, while inadequate data sovereignty protections risk exploiting vulnerable populations, such as Indigenous groups, eroding trust and perpetuating inequities (Reddy, Fox and Purohit 2019). Moreover, most health-care data remains fragmented, locked in incompatible formats, or restricted by proprietary systems, undermining the quality and scalability of AI tools (Carroll et al. 2020). Removing these barriers is critical because they limit the equitable dissemination of AI tools and compromise the ethical use of data, which is central to fostering a