Title Dirbtinis intelektas ir civilinė atsakomybė (pasirinktos problemos analizė) /
Translation of Title Artificial intelligence and civil liability.
Authors Juočeris, Kęstutis
Full Text Download
Pages 48
Abstract [eng] The Master's thesis analyses the peculiarities of the application of civil liability for damages caused by the actions of artificial intelligence (AI). For this purpose, it examines relevant legal acts, the interpretations of scholars in the legal doctrine and the conclusions established in case law. The subject of this work is revealed by examining the relationship between civil liability and AI. Proper regulation of civil liability for damage caused by AI promotes people's confidence in this technology, which leads to economic development. The paper examines the concept of AI, the principle of its operation and its types. It analyses the possibility of AI being the subject of legal relations, giving it the status of an electronic or legal person. Legal theory assumes that only human beings are participants in legal relations. Therefore, AI created by man should not be equated with human beings. As a legal subject, it would not be able to hold assets against which recovery could be sought, which would undermine the compensatory function of civil liability. Similarly, AI does not currently have consciousness and would therefore not be subject to the most important civil condition, i.e. fault. Therefore, if an AI cannot be held liable for the damage caused by its actions, the person who is responsible must be held liable for that damage. This liability arises without fault and is known as strict liability. The two types of civil liability examined in this paper are civil liability for damage caused by a major hazard and civil liability for damage caused by defective products. Artificial intelligence would, according to the current regulatory framework, correspond to the concept of a major source of danger and its operator would be an external system operator, according to the new risk management theory. The manufacturer would be liable for a defective product in the form of AI, but it may be difficult for the consumer to prove the possible defective quality of the AI, as this requires specific knowledge.
Dissertation Institution Vilniaus universitetas.
Type Master thesis
Language Lithuanian
Publication date 2022