KJCKorea
Journal Central

The development of automated driving and autonomous shipping technologies is fundamentally transforming the structure of accident causation, the allocation of civil liability, and the compensatory role of insurers. Traditionally, human conduct and negligence on the part of drivers, masters, or crew members have served as important bases for determining the causes of road and maritime accidents and allocating liability. In automated and autonomous transportation, however, artificial intelligence, automated driving and autonomous navigation systems, sensors, communication networks, data, remote operation, and cybersecurity may all contribute to the occurrence of an accident. As a result, it may become increasingly difficult to identify the ultimately liable party immediately after an accident, while, depending on the cause of the accident and the applicable legal relationships, potential liability may involve not only vehicle operators or shipowners but also manufacturers, software developers, system providers, data providers, and remote operators. This diversification of accident causation and potentially liable actors calls for a conceptual distinction between prompt compensation of victims and the ultimate allocation of liability. Compulsory motor liability insurance has developed as a victim-protection mechanism based on mandatory insurance and direct claims against insurers. Korean automobile compensation law preserves this victim-protection framework while allowing an insurer that has compensated a victim for an accident caused by a defect in an automated vehicle to seek recourse against the person legally liable for the loss. From a comparable victim-protection perspective, the United Kingdom’s Automated and Electric Vehicles Act 2018 imposes first-instance liability on insurers, subject to statutory conditions, for accidents caused by an automated vehicle while driving itself. The Automated Vehicles Act 2024 further develops the broader regulatory framework by introducing regulatory actors such as the Authorised Self-Driving Entity (ASDE) and the no-user-in-charge (NUiC) operator. Protection and Indemnity (P&I) insurance, by contrast, developed under fundamentally different historical and economic conditions. P&I Clubs provide mutual insurance against various third-party liabilities incurred by shipowners and other members and have developed a multilayered risk-distribution structure combining mutual risk sharing among members, individual Club retention, International Group pooling, and global reinsurance. This structure provides useful insights into the management of large-scale accumulation and systemic risks that may arise in autonomous shipping, particularly where a common software or algorithmic defect, communication network failure, or cyberattack affects multiple vessels. This article does not argue that compulsory motor liability insurance and P&I insurance are legally equivalent or directly interchangeable. Rather, it adopts a functional approach that identifies the distinct advantages developed by each system and examines how those functions may complement one another in the emerging technological environment of automated driving and autonomous shipping. The victim-protection and prompt-compensation functions developed in motor liability insurance may reduce delays caused by complex technological liability disputes, while subsequent recourse may reallocate the ultimate cost of liability to manufacturers, system providers, or other actors legally responsible for the accident. At the same time, the pooling and reinsurance mechanisms developed in P&I insurance may provide a useful model for distributing large-scale accumulation and systemic technological risks that exceed the capacity of an individual insurer. Accordingly, insurers’ compensatory responsibility in the age of automated driving and autonomous shipping should be reconstructed around three interrelated functions: prompt and primary compensation of victims; subsequent reallocation of the ultimate cost of liability through recourse based on the cause of the accident and the applicable legal relationships; and multilayered distribution of large-scale and systemic risks through collective risk sharing and reinsurance. Such a functional integration may prevent the uncertainties created by technological complexity from being shifted onto victims, while promoting an appropriate allocation of liability costs and a sustainable risk-distribution framework for automated driving and autonomous shipping.

This study examines the expansion and functional concentration of police investigative authority following the 2020 amendment to the Korean Criminal Procedure Act and the investigative authority reform implemented in 2021, and proposes a multi-layered accountability framework. The reform abolished the prosecutor’s general authority to direct police investigations, redefined prosecutors and judicial police officers as cooperative actors, and authorized the police to make first-stage non-referral decisions in cases where criminal suspicion is not recognized. At the same time, the law retained ex post controls including requests for supplementary investigation, corrective measures for illegality or human-rights violations, requests for reinvestigation, objections by interested parties, and judicial control over compulsory measures. In this article, “monopoly of investigative authority” does not mean that the police possess an exclusive legal monopoly over all criminal investigations. Rather, it describes the functional concentration that arises when initiation, progression, and first-stage termination of ordinary criminal investigations are combined within the same nationwide police organization. This distinction is particularly important because the Prosecution Office Act and the Act on the Organization and Operation of the Serious Crimes Investigation Agency, enacted in March 2026 and scheduled to take effect on October 2, 2026, indicate a further move toward a pluralized investigative structure. Based on current statutes and recent Korean scholarship published mainly from 2020 through 2026, this study identifies four principal issues: the effectiveness and complexity of controls over police non-referral decisions; delay and repeated movement of cases between investigative bodies; tension between the internal independence and centralization of the National Office of Investigation; and fragmentation of external and democratic oversight. The study proposes: strengthening an independent external oversight mechanism; redesigning remedies against non-referral decisions in stages; expanding judicial review centered on the courts; improving internal controls over case assignment, review, investigation periods, and audit; reinforcing both the organizational independence and public accountability of the National Office of Investigation; clarifying checks and cooperation among police, prosecution, and future serious-crime investigative bodies; and expanding investigative staffing, professional specialization, and disclosure of performance and rights-protection statistics. The central conclusion is that effective control should not restore a comprehensive hierarchical direction model. Instead, investigative independence should be combined with reason-giving, recordability, reviewability, and judicial remedies so that powerful police investigative authority can be exercised within a transparent and accountable framework.

Korea’s current autonomous police system has established a statutory foundation for treating community safety, traffic control, and the protection of socially vulnerable persons as autonomous police affairs. It has also created city and provincial autonomous police committees through which local policing demands may be reflected in regional policies. Nevertheless, the organizations and personnel responsible for carrying out autonomous police affairs remain within the national police system, while substantial authority over personnel, finance, and operational command has not been transferred. Consequently, policy-making authority, field-level implementation, and accountability for policing outcomes are divided among different institutions. This study approaches these limitations not merely as a question of allocating police functions or choosing between unified and dual organizational models, but as a problem involving the attribution of functions, organization, personnel authority, financial responsibility, command authority, and accountability for outcomes. It analyzes the current statutory and operational framework and compares the police accountability structures of Japan, Germany, the United Kingdom, and France. The comparison demonstrates that local police institutions generally assume primary responsibility for community policing, whereas the central government retains responsibility for national standards, interregional coordination, professional support, and the correction of regional disparities. Democratic control is exercised by residents and local councils, while political neutrality in individual law-enforcement activities is protected through collegiate police bodies and statutory limitations on political intervention. Based on these findings, this study proposes the phased alignment of autonomous police affairs with their implementing organizations, beginning with community-safety functions, and the establishment of city and provincial autonomous police headquarters under autonomous police committees. It further proposes the substantive transfer of personnel and ordinary command authority, the creation of stable statutory funding such as an autonomous police grant, and the legal institutionalization of joint-response and jurisdictional-dispute mechanisms between the national and autonomous police. An accountability-based autonomous police system signifies a transition from the formal decentralization of functions to responsibility-based decentralization, under which the institution exercising policing authority must explain and assume responsibility for its outcomes.