Intelligenza artificiale e responsabilità penale: tra tutela dei diritti fondamentali e nuove forme di colpevolezza.

This dissertation analyzes the impact of artificial intelligence (AI) on the foundations of criminal liability, with particular regard to the growing autonomy of intelligent systems and the tension it generates in relation to fundamental rights and the principles of personal imputation. After reconstructing the technical and regulatory evolution of AI—focusing on developments in machine learning and autonomous decision-making—the study examines the main challenges these technologies pose to traditional categories of criminal law, such as culpability, foreseeability, and imputability. The analysis highlights how algorithmic opacity, the unpredictability of self-learning systems, and the progressive detachment from human control risk undermining the foundations of criminal responsibility as shaped by constitutional principles. Special attention is devoted to advanced mobility technologies (such as autonomous vehicles and unmanned ships), which represent a paradigmatic field in which these issues emerge in all their complexity. Finally, the dissertation questions whether current legal systems are capable of integrating non-human agency without compromising the safeguards of criminal law, and proposes theoretical models for adapting the system in a way that remains consistent with constitutional principles.