Governing Algorithmic Fairness in Climate-Health Systems: A Policy Framework for Bias Mitigation in Public Sector Decision-Making

Abstract

Artificial intelligence (AI) systems are increasingly integrated into climate-health decision-making processes across the United States public sector, offering considerable capacity to model complex environmental and public health interactions while supporting resource allocation and policy planning. Despite these technological advances, the deployment of AI in this domain raises significant concerns regarding algorithmic bias, which can systematically disadvantage vulnerable populations and undermine the equity objectives of climate-health governance. The existing literature reveals a critical gap: no comprehensive policy framework has been specifically designed to govern algorithmic fairness within the climate-health nexus of American public sector decision-making. This article addresses that gap by developing an original, structured policy framework designated the Algorithmic Fairness Climate-Health Governance (AFCHG) Framework. Drawing on a systematic synthesis of peer-reviewed scholarship across AI governance, algorithmic fairness, public administration, and climate-health systems, this article employs a conceptual research design grounded in thematic analysis and policy analysis methodology. Conceptual frameworks play a critical role in emerging policy domains where empirical evidence remains fragmented, providing structured guidance that can subsequently be tested through applied research. The AFCHG Framework comprises four interdependent pillars: Policy and Legal Foundations, Governance and Accountability, Technical Bias Mitigation, and Ethics, Equity, and Inclusion. Each pillar is theorized in relation to existing scholarly evidence and grounded in the governance realities of United States public sector institutions. The article further proposes an eight-step governance flow for AI bias mitigation, supported by two summary tables that translate the framework into actionable policy guidance. The findings suggest that addressing algorithmic bias in climate-health systems requires not merely technical solutions but coordinated, multi-level governance architectures embedded within broader equity and human rights frameworks. This work contributes a theoretically rigorous and practically applicable framework for American policymakers, public administrators, and AI governance scholars.