Abstract
<jats:p>The growing ubiquity of artificial intelligence (AI) across domains of high social, economic, and political criticality demands a re-evaluation of the ethical underpinnings of responsibility. Conventional frameworks, anchored in notions of consciousness and phenomenal intentionality, are insufficient for contending with the nature of contemporary autonomous systems, generating a hazardous 'responsibility gap.' This paper advances the thesis that a robust and applicable AI ethics must be founded upon a functional redefinition of agency, disassociated from internal mental states. Building on Luciano Floridi's ontology of information, a conception of artificial agency is put forward as an interactive, autonomous, and adaptive capability within the infosphere. Such a perspective allows for the construction of a coherent normative framework for delegated morality and the ascription of distributed, faultless responsibility. Uniting philosophical analysis with insights from empirical moral psychology, it is contended that this approach not only clarifies unresolved normative tensions concerning artificial agency but also offers operative criteria for orienting the design and governance of digital technologies that can re-ontologize the epistemic and ethical structures of our contemporary reality. </jats:p>