Blaming Galileo: Liability for Damage Caused by GNSS Enabled Autonomous Systems

Abstract

GNSS offer solutions for many sectors, from road traffic, aviation, emergency-response services, civil engineering and agriculture. Due to the latest technological developments, GNSS, including Galileo, are also being integrated as an essential component of AI systems with various automation levels, such as self-driving vehicles, drones, lane keeping systems on highways etc. Despite their numerous benefits, GNSS are not risk-free. Even though it is unlikely that a loss of signal will lead to an accident caused by an AI system, this scenario cannot be totally ignored. Recent incidents revealed a series of vulnerabilities that need to be addressed before more AI systems using GNSS signals can become active participants in our societies. In this context, it becomes clear that the most pressing issue is the one related to liability: who will be liable in case an accident is caused by an AI system due to an absent or inaccurate GNSS signal at a critical point during navigation? Taking into consideration the debates concerning Galileo’s potential acceptance of liability, this paper investigates if international space law is able to prevent potential liability gaps, thus avoiding situations where incidents occur and liability cannot be attributed.

Keywords: artificial intelligence, liability, GNSS, Galileo