Normative Principles for the Legal Regulation of Artificial Intelligence: Human Dignity and Non-Delegable Agency

The proliferation of increasingly sophisticated artificial intelligence systems has produced a critical inflection point in the legal understanding of the relationship between humanity and technology. In this context, the article proposes a philosophical legal framework that, drawing on a radical ontological distinction between human beings and artificial systems, identifies normative principles aimed at regulating their development without compromising human dignity. Grounded in the tradition of moderate realism, the analysis argues that artificial systems, lacking substantial unity, conscious interiority, and genuine intentionality, cannot be regarded as ontological subjects equivalent to human persons. This separation imposes non-negotiable normative limits on the design and use of artificial intelligence and requires regulatory approaches that preserve spheres of inalienable human agency. The study concludes that the exclusivity of human moral judgment, the imposition of boundaries on algorithmic delegation, and the adoption of technological subsidiarity should guide regulatory frameworks intended to safeguard domains reserved for human agency within increasingly automated environments.