Artificial Intelligence and Civil Liability
Jānis Kārkliņš
Abstract
Although modern innovations already provide an artificial intelligence-driven devices or software’s capability to function fully autonomously without the involvement of any person, the issue of the specificity of the application of civil liability is, inter alia, weighed at the European Union level. This is linked to the current legal framework of the European Union and its Member States, which has been developed and adopted over a period of time when the operation of devices or software has requested at least an indirect involvement of a person in order to execute a specific task. In contrast, contemporary technological achievements in the design of artificial intelligence provide the ability of an artificial intelligence-driven device and software to take independent decisions regarding the conduct of a particular activity, or, on the contrary, on abstaining from carrying out a particular activity, without any person’s involvement. In such circumstances, there is a need to review and reassess the content of the existing regulation with regard to the application of civil liability both in the fault liability and in the model of strict liability (including product liability), and to establish certainty in the conceptual understanding of artificial intelligence and its isolation from simple process automation.