RESPONSABILITATEA PENALĂ ÎN ERA INTELIGENŢEI ARTIFICIALE, DE LA AUTORII UMANI LA AGENŢI AUTONOMI
CRIMINAL LIABILITY IN THE ERA OF ARTIFICIAL INTELLIGENCE, FROM HUMAN PERPETRATORS TO AUTONOMOUS AGENTS

Summary/Abstract: In the era of artificial intelligence (AI), criminal responsibility takes on new dimensions as autonomous agents take on complex tasks, acting independently of direct human control. The article explores the challenges posed by criminal liability in cases involving the actions of these agents, highlighting the difficulties in applying traditional concepts of guilt and causality, which are essential to classical criminal law. Traditionally, criminal responsibility was linked to actions committed by individuals or legal entities, where intent and culpability play a central role. However, AI complicates this legal framework, as autonomous agents lack consciousness, intent, or guilt, making it difficult to assign responsibility in cases such as accidents caused by autonomous vehicles or errors made by medical algorithms. The article discusses the solutions proposed by legal doctrine and case law, such as indirect liability of developers and companies that implement AI, as well as the need to create new legal frameworks to regulate autonomous agents. Various international legislative approaches are analyzed, and the creation of new categories of civil or criminal liability is proposed to account for the specific nature of autonomous agents. Additionally, the ethical implications of AI agents' responsibility are debated, along with the future challenges that emerging technologies pose to the legal system.

• Details
• Contents