A Tale of Two Anthropocentric Abstractions: On the Concept of the “Human” in AI and Environmental Ethics

This paper examines how anthropocentric assumptions shape the normative frameworks of both AI ethics and environmental ethics, arguing that each domain is dependent on a distinct form of anthropocentric abstraction that is increasingly conceptually unstable. In environmental ethics, anthropocentrism is frequently critiqued as an ontological and moral limitation that obscures nonhuman entities and downplays ecological interdependence. By contrast, the European approach to AI ethics – grounded in principles of trustworthy AI, transparency, risk-based regulation, and the protection of fundamental rights – tends to reaffirm an anthropocentric framework, prioritizing human values, rights, and responsibilities in addressing algorithmic opacity, bias, and environmental impacts, while largely overlooking a critical perspective on the shortcomings and harms perpetrated in the name of a misguided or overly narrow understanding of human-centeredness. Drawing on metaethical resources, the paper disentangles the implicit conceptions of the “human” operative in each domain. It argues that both domains grapple with abstract conceptions of the human, either as the telos of nature – occupying a privileged, central role as if all of creation were oriented toward human flourishing – or as fundamentally defined by moral agency, marked by the capacity of being morally responsible for actions, omissions, consequences, and emotions. These tensions highlight the need for an approach that critically engages with the idea of human-centeredeness, while resisting reliance on a merely assumed – and often contested – appeal to “the human”. By clarifying how different anthropocentric models operate across ethical contexts, the paper contributes to a more coherent and critically human-centered ethical framework for addressing both technological and ecological challenges.