The Application of Artificial Intelligence in the Processes of Evaluation and Quantification of Physical impairment: Methodological, Ethical, and Legal Perspectives

This doctoral thesis examines the application of artificial intelligence (AI) in the evaluation and quantification of personal injury within the framework of civil liability, with particular reference to medico-legal assessment and the private insurance sector. The research adopts an interdisciplinary approach that combines civil law analysis, forensic medicine, and the European regulatory framework on artificial intelligence and personal data protection. The study focuses on the medico-legal evaluation of biological damage, which represents a central step in the determination of compensation for personal injury. This evaluative process traditionally relies on expert medico-legal judgment aimed at translating clinical evidence concerning the impairment of psychophysical integrity into legally relevant parameters capable of guiding the quantification of damages. The growing interest in AI-based tools for damage assessment raises important methodological, ethical, and legal questions regarding transparency, reliability, and the preservation of individualized evaluation. Particular attention is devoted to the role of health data used for the training of algorithmic models, the relationship between automated outputs and human professional judgment, and the potential impact of algorithmic systems on fundamental rights, including the protection of personal data and the principle of personalized compensation. The thesis argues that the primary challenges related to the use of AI in personal injury assessment do not lie solely in the technical accuracy of algorithms, but rather in the legal legitimacy of the data processing practices underlying algorithmic systems and in the risk of transforming medico-legal evaluation into a standardized computational procedure. In this perspective, the study distinguishes between AI systems used as decision-support tools and those functioning as decision- substituting mechanisms. The research concludes that artificial intelligence may legitimately support medico-legal evaluation only when effective human oversight and professional judgment remain central to the decision-making process, ensuring consistency with the principles of civil liability and the protection of human dignity.