Algorithmic liability

Abstract

The idea of smart robots has always been present in popular culture, mainly and un­til recently as science fiction, whereas now it is deemed a reality. Using the deductive method and bibliographical review, the present work advocates for the possibility of assigning to ar­tificial intelligence machine an ‘electronic per­sonality’ (‘e-personality’, as mentioned in this study), which would enable the recognition of their liability in cases where third parties have been harmed without human interference, while also analyzing the alternative to attribute liabili­ty to persons. The paper gives a take on different theories of civil liability, foreign court decisions and legislation, concluding that the present legislative framework is not enough to deal with the issue presented herein.