Governing Sustainable Artificial Intelligence: A Green AI Regulatory Framework Based on Responsible Innovation

Abstract

This study aims to examine how sustainability is incorporated into AI governance and to address the resulting policy gaps.  Using an analytical-comparative normative policy approach, this study analyzes AI regulatory frameworks across six jurisdictions: the European Union, the United States, China, Japan, South Korea, and Singapore. The analysis is guided by the Responsible Innovation (RI) framework, which serves as a conceptual lens for assessing the integration of sustainability across four dimensions: anticipation, reflexivity, inclusion, and responsiveness. The findings reveal that current AI governance frameworks lack explicit mechanisms for environmental impact assessment, energy and emissions reporting, and adaptive sustainability regulation. To address these gaps, this study proposes a Green AI regulatory framework consisting of three interconnected layers: sustainability assessment, governance and compliance, and technical and innovation. This study contributes theoretically by bridging AI governance with the RI framework to operationalize sustainability in regulatory design. Practically, it provides conceptual guidance for policymakers and regulators to integrate environmental sustainability, particularly in energy consumption, carbon emissions, and resource use, into AI governance. The proposed framework supports the development of more transparent, accountable, and environmentally sustainable AI systems aligned with global sustainability goals.