Sustainable AI and the Imperative of Global Distributive Justice by Inês Neves
This guest blog was authored by DISC Virtual Visiting Fellow Inês Neves, PhD in Law and Invited Assistant Professor at the Faculty of Law of the University of Porto, Portugal. In this post, Inês explores the role of Law in addressing the unequal distribution of the environmental and social impacts associated with AI infrastructures and value chains. Drawing on principles of sustainable development and global distributive justice, she examines the potential of extraterritorial regulatory approaches and a model of shared responsibility among States, companies, international organisations, and civil society to advance more sustainable and equitable AI governance.
Artificial Intelligence (‘AI’) has emerged as one of the defining technologies of the twenty-first century (IEA, 2025). Increasingly, it shapes economic activity, public administration and social interaction, while recent developments in generative AI have expanded its capacity to influence activities long regarded as inherently human. The questions raised by AI therefore extend beyond technological innovation and concern the conditions under which contemporary societies organise production, knowledge and collective life.
The apparent immateriality of AI conceals a physical and social infrastructure dependent upon data, computational power, energy, water and critical raw materials. The expansion of data centres and AI-related infrastructures has drawn increasing attention to the environmental implications of digitalisation, including energy consumption, water use, greenhouse gas emissions, resource extraction and labour conditions across global supply chains (see, among others, Chouksey et al. 2026; IEA, 2025; Winter and Züger, 2025; van der Ven et al., 2024; Xiao et al., 2025). International organisations and standard-setting initiatives have likewise highlighted the need for common methodologies, reporting frameworks and accountability mechanisms capable of assessing AI’s environmental footprint (Coalition for Sustainable AI, 2026; ITU, 2026; UNEP, 2024).
Recent scholarship has also drawn attention to the distributive implications of AI in the context of climate action (see, among others, Jian, Xian and Chen, 2025; Li et al., 2024; Machen and Pearce, 2025; UNEP, 2024b). Environmental and social impacts associated with AI infrastructures are not distributed evenly. More demanding environmental standards in certain jurisdictions may create incentives for the relocation of environmentally intensive activities to jurisdictions where environmental regulation, labour protection or enforcement mechanisms are less stringent, while economic and technological benefits tend to concentrate in particular regions and among particular actors.
While “most literature on the topic is concentrated in computer science, engineering, and natural sciences” (Kudina et al., 2026), the sustainability of AI also presents important normative and legal questions. The concentration of technological and economic benefits in certain regions, coupled with the displacement of environmental and social burdens to others, raises a question of global distributive justice and invites examination through the principle of sustainable development and its underlying commitments to equity and fairness. It also raises a legal question concerning the extent to which legal systems may legitimately respond to situations in which the benefits and burdens associated with technological development are distributed across different jurisdictions and populations. As Coeckelbergh and Sætra observe, “AI and climate change challenge both national and global governance structures, and we should be open to considering different forms of governance at all levels.” (Coeckelbergh and Sætra, 2023).
The research proceeds from the premise that sustainable development possesses both a temporal and a spatial dimension. Alongside concerns relating to present and future generations, it encompasses questions of justice between States, regions and communities situated differently within the global economy. From this perspective, the environmental and social impacts associated with AI infrastructures and value chains raise questions concerning legal responsibility and the allocation of regulatory obligations.
Against this background, the analysis examines the role of constitutional law, fundamental rights and public regulation in addressing the distributive consequences of AI-related environmental impacts. The central hypothesis is that constitutional orders, as well as constitutionalised supranational entities such as the European Union, provide a legitimate basis for the adoption of legal obligations requiring business actors subject to their jurisdiction to comply with minimum standards of environmental responsibility and resource efficiency in relation to the development and deployment of AI systems.
Particular attention is devoted to the possibility of attaching such obligations to economic actors irrespective of the geographical location in which environmental impacts materialise. The argument advanced is that where regulatory authority rests upon a sufficiently robust jurisdictional connection between the regulator and the regulated entity, the extraterritorial effects that may follow do not constitute an illegitimate interference with the sovereignty of third States, but rather a dimension of “sovereignty as responsibility” (Huttunen, 2024) or the reconfiguration of the “the state duty to protect as a duty of transnational cooperative governance” (Khan, 2026). The obligations in question apply to companies subject to the jurisdiction of the regulating authority rather than to foreign entities as such. Their purpose is to address the externalisation of environmental costs and to ensure that sustainability obligations accompany those actors who derive benefit from the relevant economic activity.
The relevance of this approach becomes particularly apparent where regulatory asymmetries create incentives for environmental displacement. The question then becomes whether legal systems committed to sustainable development may legitimately require economic actors subject to their jurisdiction to comply with higher standards irrespective of the location of their operations.
The position defended is that they may. A conception of sustainable development that incorporates commitments to equity and fairness sits uneasily with patterns of technological development that allow environmental harms to be displaced while benefits remain concentrated elsewhere. Sustainable AI therefore presupposes a commitment to global distributive justice as a condition of its legitimacy. Such a commitment supports legal frameworks capable of ensuring that responsibility accompanies those actors whose activities contribute to the environmental and social costs associated with AI infrastructures and value chains.
Building on the United Nations’ Pact for the Future and Global Digital Compact message regarding the need to ensure that digital technologies contribute to sustainable development and are governed in ways that promote inclusion, equity and shared prosperity (United Nations, 2024), this perspective recognizes a role for Law in ensuring a framework of shared responsibility involving States, international organisations, economic actors and civil society. The obligations borne by each may differ in scope and intensity as a matter of proportionate legal design. The underlying concern remains the same: ensuring that technological development takes place within a legal framework capable of addressing the distributive consequences of its environmental and social footprint.
