1.1 Thesis Topic, Research Questions, and Overview
As the technological developments of Maritime Autonomous Surface Ships (MASS) and artificial intelligence (AI) in shipping continue, the need to address new legal issues arises. Issues regarding liability in future accidents involving MASS have been frequently discussed in international legal literature in recent years.(1) Inter alia in Béatrice Schütte, 'Damage caused by autonomous ships: towards regulation for civil liability in EU waters?' in Ellen J Eftestøl, Anu Bask and Maximilian Huemers (eds), Towards a Zero-Emissions and Digitalized Transport Sector: Law, Regulation, and Logistics (1st edn, Edward Elgar Publishing 2024); Bariş Soyer and Andrew Tettenborn, 'Autonomous Ships and Private Law Issues' in Bariş Soyer and Andrew Tettenborn (eds), Artificial Intelligence and Autonomous Shipping: Developing the International Legal Framework (1st edn, Hart Publishing 2021); Henrik Ringbom, Erik Røsæg and Trond Solvang (eds), Autonomous Ships and the Law (1st edn, Routledge 2021) Part III. Meanwhile, there has been less discussion on operational and regulatory requirements for these ships.
The Artificial Intelligence Act (AIA) was adopted by the EU in August 2024.(2) Regulation (EU) 2024/1689 (AIA). Its requirements for "high-risk AI systems" address issues such as transparency, human oversight measures, and cybersecurity.(3) The AIA Articles 13, 14 and 15, respectively. These are central considerations in the development of MASS, particularly in relation to autonomous navigation systems (ANS).(4) Regarding transparency, see Andreas Nygard Madsen and Tae Eun Kim, 'A state-of-the-art review of AI decision transparency for autonomous shipping' (2024) 8 J. int. marit. saf. Regarding human oversight measures, see J (Hans) van den Broek, J R (Jaco) Griffioen and M (Monique) van der Drift, 'Meaningful Human Control in Autonomous Shipping: An Overview' (2020) 929 Mater. Sci. Eng. 012008, Section 3. Regarding cybersecurity, see Nimra Tabish and Tsai Chaur-Luh, 'Maritime Autonomous Surface Ships: A Review of Cybersecurity Challenges, Countermeasures, and Future Perspectives' (2024) 12 IEEE 17114.
Against this background, the topic of this thesis is the significance of the requirements for high-risk AI systems in the AIA in the context of ANS. The topic will be examined through the lens of two research questions. First, in Chapter 3, I examine the substantive scope of the requirements and consider whether they may apply to ANS deployed in commercial cargo shipping. Second, in Chapter 4, I presume that the requirements do not apply to such ANS and consider whether the requirements may nevertheless have an impact on other regulations governing ANS. For the purpose of this thesis, I refer to this as the requirements having an "indirect impact".
Before I address these research questions, Chapter 2 sets out their actual and legal premises. This includes a conceptual clarification of ANS, and an introduction to the interplay between the EU and the International Maritime Organization (IMO) in the context of regulating MASS and ANS. Finally, in Chapter 5, I present conclusions to the research questions, and reflect on possible approaches that the EU may adopt with respect to regulating ANS in the near future.