Punitive intellectual property (IP) frameworks and inadequate data sovereignty protections are significant barriers to equitable artificial intelligence 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.
