Creative Commons

This dissertation has a goal of approaching the master of liability in cases where there is Artificial Intelligence operation involved. Specifically, the following paper is tackling the rapid integration of AI systems into diverse sectors of public and private life, from transportation to healthcare, which presents immense opportunities alongside critical ethical and legal challenges. Furthermore, this dissertation delves into the complex topic of liability attribution surrounding such AI systems, focusing on its impact on individuals and society. The research begins by defining AI and outlining its unique risk profile compared to traditional systems. It then explores the current legal landscape, commenting on EU and UN approaches as well as State national law approaches, highlighting the problematic that arises in addressing the nuances of AIinduced harm, particularly due to the "black box" nature of algorithms and the potential for data bias. The main part of the dissertation focuses dissecting the multifaceted roles of various stakeholders within the AI lifecycle – developers, deployers, and end users. By analyzing their responsibilities and potential negligence, the research untangles the intricate web of accountability when harm manifests. Leveraging real-world cases and emerging legal frameworks, the dissertation proposes a frame for navigating the mostly uncertain and opaque problematic of AI liability. It advocates for clear legal standards, transparency and explainability in AI models, and emphasizes robust risk management protocols. Ultimately, this dissertation strives to pave the way for responsible development and deployment of AI systems, ensuring trust, fairness, and safety in a world increasingly shaped by intelligent machines.