Joshua Gellers

Efforts to promote “Sustainable AI” must confront the economic realities shaping corporate behavior rather than rely on aspirational or purely voluntary approaches. A central tension defines this agenda: although AI development is accelerating amid urgent climate concerns, companies lack sufficient incentives to disclose or reduce the environmental impacts of their systems. Three key microeconomic barriers help explain this reluctance. First, a first-mover disadvantage discourages firms from voluntarily revealing environmental data that competitors could exploit. Second, intense competitive pressures, especially within a global race for AI dominance, prioritize speed, innovation, and market share over sustainability. Third, fiduciary duties to maximize shareholder value, coupled with declining support for environmental, social, and governance (ESG) initiatives, further weaken motivations for environmental transparency. Insights from political science research on voluntary environmental governance offer a more pragmatic path forward. This literature suggests that sustainability initiatives must align with market incentives to succeed. Effective strategies include establishing standards across key points in the AI supply chain, developing credible certification systems (“green clubs”) that balance rigor with accessibility, and relying on multi-stakeholder or third-party governance structures to enhance legitimacy. Hybrid models that combine industry participation with civil society oversight appear particularly promising, as they can generate both demand for and trust in environmental certifications. Ultimately, truly Sustainable AI will remain elusive without reforms that integrate economic incentives with governance mechanisms. Designing systems that make environmental disclosure compatible with corporate interests is essential to bridging the gap between ecological responsibility and technological advancement.

Philosophers love a hierarchy. Nothing seems to fit the needs and desires of the moral imagination more than the promise of a clean, orderly ladder of moral worth with (not surprisingly) humans at the top, a few “higher” animals somewhere beneath, plants and rocks at the bottom, and now, far outside the frame, artificial agents politely waiting their turn. In addition, it is in the context of AI that this impulse to police the moral boundary has returned with renewed urgency. Many commentators, including Adrianna de Ruiter in her recently published essay ‘Dangerous Liaisons’ (2025), warn that relational approaches to AI moral status are dangerous, misleading, or lacking philosophical rigor. They insist that we must firmly anchor moral considerability in some intrinsic, scientifically respectable property (i.e., sentience, consciousness, phenomenal experience, and vulnerability) and that anything less—especially when it comes to AI—is dangerous.

Jay Odenbaugh offers a wildlife management solution to the dilemma involving competing owl species. However, the topic of rights is curiously absent from this proposal. In this response, I argue that killing owls is morally and legally wrong whether approached from the perspective of animal rights or rights of nature. I contend that no matter which approach is taken, the conflicting state of various owl populations does not necessitate a duty to kill some portion of them. Instead, we should translate our grief over our moral failings into political action in the hopes of securing a better future for all.

Artificial intelligence (AI) is developing rapidly in an era in which the urgent need to address climate change is being confronted by anti-regulatory headwinds. These opposing forces complicate the pursuit of Sustainable AI, an ambitious goal that requires motivating private actors to reduce their AI-driven environmental impacts at the same time that the environmental consequences of AI remain highly uncertain. In this essay, I present a productive way to overcome this conundrum. Acknowledging the real issues that stand in the way of companies disclosing environmental performance and drawing practical insights from scholarship on voluntary environmental governance, I propose a Sustainable AI that shuns wishful thinking in favor of a healthy dose of pragmatism.

This Commentary offers a critique of an eco-relational approach in robot ethics, highlighting the importance of articulating an ecologically-sensitive ethical orientation that incorporates the entire more-than-human world, including technological entities like forms of artificial intelligence. While the eco-relational approach enhances our understanding of the complex way in which morally significant properties operate on a phenomenological level, it is not without its flaws. In particular, this perspective focuses on ethical concepts when it needs to be rooted in ethical systems, misrepresents the extant literature on relational approaches to moral status, and is insufficiently ecological in a way that would bring machines into the moral universe by way of their location in and importance to the larger environment. At a time when the legal status of robots is being decided before moral questions are settled, it remains important to get the contours of relevant debates right so that the robot ethics discourse brings us closer to developing practical solutions for ethical dilemmas emerging in our increasingly technological age.

Bringing a unique perspective to the burgeoning ethical and legal issues surrounding the presence of artificial intelligence in our daily lives, the book uses theory and practice on animal rights and the rights of nature to assess the status of robots. Through extensive philosophical and legal analyses, the book explores how rights can be applied to nonhuman entities. This task is completed by developing a framework useful for determining the kinds of personhood for which a nonhuman entity might be eligible, and a critical environmental ethic that extends moral and legal consideration to nonhumans. The framework and ethic are then applied to two hypothetical situations involving real-world technology—animal-like robot companions and humanoid sex robots. Additionally, the book approaches the subject from multiple perspectives, providing a comparative study of legal cases on animal rights and the rights of nature from around the world and insights from structured interviews with leading experts in the field of robotics. Ending with a call to rethink the concept of rights in the Anthropocene, suggestions for further research are made. An essential read for scholars and students interested in robot, animal and environmental law, as well as those interested in technology more generally, the book is a ground-breaking study of an increasingly relevant topic, as robots become ubiquitous in modern society.

Sing C. Chew’s latest book, Ecology, Artificial Intelligence, and Virtual Reality: Life in the Digital Dark Ages, complements his trio of monographs on the history of environmental degradation and social change. His timing could not be better, as books examining the intersection of environmental crisis and emerging technologies have quickly become a cottage industry. This burgeoning literature generally consists of two camps – techno-optimists (such as those contributing to Fei Fang et al.’s (2019) edited volume Artificial Intelligence and Conservation) and techno-pessimists (exemplified by Peter Dauvergne’s (2020) AI in the Wild). Chew’s title, despite its reference to the Dark Ages, fits comfortably in the middle of these opposing perspectives.

While scholars have expended great effort analyzing environmental discourse and applying a critical lens to environmental law, scant work has used critical discourse analysis to study environmental law. This is surprising given the rising prominence of CDA and the continued development of critical environmental law scholarship. The present article seeks to correct for this oversight by highlighting the particularities of environmental law which compel the use of CDA, and outlining a method by which social science researchers can use CDA to understand the role of power in the domain of environmental law.

Leading minds in the popular and scholarly discourse on artificial intelligence (AI) ethics tend to issue strong claims with unshakeable confidence—AI is just a tool, AI does not “think” or “learn,” humans can just stop anthropomorphizing AI, etc. Oddly, otherwise brilliant scholars and commentators seemingly hit an intellectual wall when it comes to contemplating the ontological and moral status of this rapidly advancing technology. The staleness of this discursive state of affairs is all the more surprising given the commitment of many AI experts to diversity, social justice, and environmental sustainability, values that should counsel against expressions of certainty and welcome rather than reject a plurality of perspectives. Thus far, few attempts have been made to formally expose these totalizing tendencies and ideological contradictions with an eye towards laying the groundwork for more inclusive conversations about the ethical implications of AI.