This paper suggests that AI regulation needs a shift from trustworthiness to sustainability. With the carbon footprint of large generative AI models like ChatGPT or GPT-4 adding urgency to this goal, the paper develops a roadmap to make AI, and technology more broadly, environmentally sustainable. It explores two key dimensions: legal instruments to make AI greener; and methods to render AI regulation more sustainable. Concerning the former, transparency mechanisms, such as the disclosure of the GHG footprint under Article 11 AI Act, could be a first step. However, given the well-known limitations of disclosure, regulation needs to go beyond transparency. Hence, I propose a mix of co-regulation strategies; sustainability by design; restrictions on training data; and consumption caps. This regulatory toolkit may then, in a second step, serve as a blueprint for other information technologies and infrastructures facing significant sustainability challenges due to their high GHG emissions, e.g.: blockchain; metaverse applications; and data centers. The second dimension consists in efforts to render AI regulation, and by implication the law itself, more sustainable. Certain rights we have come to take for granted, such as the right to erasure (Article 17 GDPR), may have to be limited due to sustainability considerations. For example, the subjective right to erasure, in some situations, has to be balanced against the collective interest in mitigating climate change. The paper formulates guidelines to strike this balance equitably, discusses specific use cases, and identifies doctrinal legal methods for incorporating such a "sustainability limitation" into existing (e.g., Art. 17(3) GDPR) and future law (e.g., AI Act). Ultimately, law, computer science and sustainability studies need to team up to effectively address the dual large-scale transformations of digitization and sustainability.

翻译：本文提出人工智能监管需要从可信性转向可持续性。随着ChatGPT或GPT-4等大型生成式人工智能模型的碳足迹日益加剧这一目标的紧迫性，本文制定了一条使人工智能及更广泛的技术实现环境可持续的发展路线图。文章探讨了两个关键维度：使人工智能更环保的法律工具，以及使人工智能监管更具可持续性的方法。在前者方面，透明度机制（例如根据《人工智能法案》第11条披露温室气体足迹）可作为第一步。然而，鉴于信息披露的已知局限性，监管需超越透明度范畴。因此，本文提出混合型共同监管策略、可持续设计原则、训练数据限制以及消费上限的组合方案。这一监管工具箱随后可作为其他面临高温室气体排放重大可持续性挑战的信息技术和基础设施（例如区块链、元宇宙应用、数据中心）的范本。第二个维度涉及使人工智能监管（进而使法律本身）更具可持续性的努力。某些我们习以为常的权利（如《通用数据保护条例》第17条规定的删除权）可能因可持续性考量而需受到限制。例如，在某些情况下，删除权的主观权利须与减缓气候变化的集体利益相平衡。本文制定了公平实现这种平衡的指导原则，讨论了具体应用场景，并确定了将这种"可持续性限制"纳入现有法律（如《通用数据保护条例》第17条第3款）和未来法律（如《人工智能法案》）的学说性法律方法。最终，法律、计算机科学与可持续性研究需协同合作，以有效应对数字化与可持续性的双重大规模转型。