Ethical Governance of Bioinformatics and Genomic AI Systems: From Compliance to Institutional Legitimacy

Abstract

Bioinformatics and genomic artificial intelligence (AI) systems are increasingly embedded in clinical and organizational decision-making. However, ethical analysis in this area remains dominated by regulatory compliance, privacy protection, and technical control. Although these approaches are necessary, they are analytically insufficient because they do not adequately explain how ethical governance is enacted within organizations, how technological practices mediate governance, or why formally compliant systems may still fail to achieve institutional acceptance. This article develops a conceptual framework through interdisciplinary analysis of biomedical informatics governance scholarship, socio-technical systems theory, organizational theory, and legitimacy theory. Its aim is not merely to synthesize these literatures, but to address an explanatory gap concerning how ethical governance is organizationally produced and how its effectiveness should be evaluated. The proposed Culture–Technology–Ethics–Legitimacy (CTEL) framework links four interdependent domains: organizational culture, bioinformatics technology practices, ethical governance practices, and legitimacy outcomes. The framework makes three conceptual contributions. First, it redefines ethical governance as a cross-level socio-technical process rather than a compliance endpoint. Second, it conceptualizes technology as an ethical mediator through which organizational norms are translated into operational practice. Third, it positions legitimacy as the key downstream outcome of governance and distinguishes scientific, clinical, social, and moral legitimacy as analytically distinct but interacting forms of institutional acceptance. By repositioning legitimacy as the central evaluative outcome of ethical governance, the framework advances biomedical informatics ethics beyond compliance-driven approaches and offers a stronger explanatory account of why some bioinformatics and genomic AI systems become institutionally trusted while others do not. The framework therefore offers a distinct conceptual contribution and a foundation for future empirical investigation