With the advancement of artificial intelligence technologies and their increasing role in independent decision-making, the issue of granting legal capacity to these systems has become a major legal challenge. This research, using an analytical-comparative method, examines the conditions and consequences of this action, focusing its main analysis on civil liability and the possibility of insuring AI activities. Traditional legal principles limit legal capacity to natural and legal persons (such as corporations) who possess independent will and intent. However, the progress of intelligent systems, especially autonomous ones, raises the question of whether, inspired by the concept of legal personality, a similar capacity can be considered for AI so that it can hold rights and duties and be directly held responsible for damages caused by its actions. This is particularly crucial in the field of civil liability; because in common practice, liability ultimately falls back on the manufacturers, owners, or users of the systems, which seems inadequate or unfair as AI's autonomy increases. The research findings indicate that granting limited legal capacity to AI, although complex, is feasible and can lead to a more equitable distribution of liability and more effective compensation for damages. This requires redefining legal concepts such as will and intent, as well as establishing regulatory frameworks and registration systems. In the field of insurance, this approach could lead to the creation of "specific liability insurance for AI," where the intelligent system has insurance coverage as an independent insured subject. Ultimately, realizing this idea necessitates a fundamental revision of current laws and the development of new regulations that recognize a novel legal personality for highly advanced intelligent systems, clearly defining the scope of their rights and duties, including liability for damages and insurance mechanisms.