The Governance Vacuum in Medical Device AI: Toward an Equitable and Accountable Framework

Background: The rapid adoption of artificial intelligence (AI) in the medical device sector has outpaced the development of regulatory frameworks capable of ensuring fairness, safety, and accountability. This has created a governance vacuum where systemic bias, flawed proxy variables, and emergent risks to patient safety persist unaddressed. Existing models often default to ethical generalities without mechanisms for operational enforcement.Methods: To address this, we conduct a comprehensive policy analysis drawing on global regulatory precedents, illustrative case studies, and three critical frameworks: FUTURE-AI, the Health AI Readiness Assessment (HAIRA), and Public Health Critical Race Praxis (PHCRP).Findings: We show that governance failures in medical AI are often structural, not incidental, but rooted in insufficient institutional readiness and an absence of equity-centered design. In response, we propose a novel, integrated, lifecycle-oriented governance framework that expands upon existing models. Central to this framework is the concept of enforceable equity: the translation of ethical principles into auditable standards and decision-making protocols across the AI lifecycle.Interpretation: This framework confronts the inherent friction between ethical ambition and practical implementation, offering a practical blueprint for aligning innovation with accountability. By embedding enforceable equity into both organizational processes and technical evaluation, it provides a path toward trustworthy, fair, and resilient AI in the medical device industry.