AI and climate: an ethical sustainability framework for balancing risks and responsibilities

Abstract Artificial intelligence has become a central topic in multidisciplinary debates, particularly in relation to the climate crisis. This paper critically examines the dual role of AI as both a contributor to environmental challenges and a potential tool for climate mitigation and adaptation. Although AI-based technologies, especially generative models, consume significant energy and contribute to CO 2 emissions, they also offer innovative solutions for energy optimization, disaster prediction, and sustainable resource management. The article takes a critical account of the current risk-based regulatory approach, exemplified by the EU AI Act ((AI 2024)), arguing that it prioritizes harm prevention over proactive responsibility. As a form of integration, it proposes the concept of a Sustainability Alignment Tax (SAT) as a conceptual framework to balance AI development with environmental and socio-economic sustainability. The SAT encourages stakeholders to weigh the benefits of AI technologies against their environmental and social costs, fostering a more ethical and responsible approach to AI deployment. By integrating principles of responsibility and sustainability, the document advocates for a regulatory shift that aligns AI innovation with global climate goals, ensuring that the benefits of AI are equitably distributed although minimizing its socio-ecological footprint.