Inteligência artificial e pessoalidade

Abstract

This paper aims to investigate if Law should regard autonomous systems based on artificial intelligence algorithms as persons, whether in the formal or substantial sense of the term, and to whom liability should be attributed in case of any damage caused by them. In order to answer these questions, it will be first investigated if such autonomous agents do possess the subjective states essential for intrinsic intentionality as a condition to the dimension of agency, using John Searle’s conception of consciousness as theoretical framework. Following, the reasons that inform the personification of legal entities will be analyzed in order to assess if they can also be applied to artificial intelligence systems or if it would represent a manifestation of the phenomena Rodotà named as “expropriation of subjectivity”. The research concludes that autonomous artificial agents should not be qualified as persons by Law, and that the liability for any damages caused by them should lie on those who made the technology available to the public following a model of aggravated strict liability.