Liability and Insurance in the Era of Autonomous Vehicles

Chun Ji-yeon, Researcher at the Insurance Research Institute [Photo=Insurance Research Institute]

Who is liable for accidents involving autonomous vehicles? In traditional car accidents, the answer is relatively straightforward: liability is determined based on the driver's fault, and damages are compensated through auto insurance. However, in the era of autonomous vehicles, the responsibility for accidents is no longer limited to the driver alone. As control shifts to the autonomous driving system (ADS), various parties—including vehicle owners, manufacturers, software developers, operators, telecommunications providers, and road infrastructure managers—can be considered contributors to the cause of an accident. The commercialization of autonomous vehicles demands not only a change in driving methods but also a transformation in the liability framework and insurance systems.

The global autonomous driving market is rapidly expanding. In the United States, companies like Waymo are accumulating extensive operational data through their robo-taxi services, while Tesla is advancing its Full Self-Driving (FSD) technology. In China, the government is promoting infrastructure development and regulatory easing to expand its driverless taxi services. South Korea is also working to establish regulations and infrastructure for the commercialization of Level 4 autonomous vehicles by 2027, including the operation of nationwide pilot zones and the creation of autonomous driving demonstration cities.

To effectively respond to the commercialization of autonomous vehicles, discussions regarding the liability framework are essential. Given that most compensation for damages occurs through auto insurance, it is crucial to reform the insurance system. Understanding the stages of autonomous driving technology is a necessary first step. Currently, advanced driver-assistance systems (ADAS) are at Levels 1 and 2, where the human driver remains in control. In contrast, from Level 3 onward, the autonomous driving system takes over driving responsibilities, and at Level 4 and above, full autonomy is possible without driver intervention under specific conditions (Operational Design Domain, ODD). Ultimately, at Levels 4 and above, the key issue becomes determining who is liable when an accident occurs.

In response, major countries are establishing new liability frameworks. The United Kingdom, Germany, and Japan are defining the obligations and responsibilities of manufacturers and operators during autonomous operations and are reforming related regulations. Observing these developments suggests that the focus of liability is shifting from individual drivers to system operators in the autonomous driving era. However, most discussions are centered around corporate-operated services like robo-taxis and autonomous shuttles, indicating that further examination is needed regarding the liability framework for privately owned autonomous vehicles.

From the insurance industry's perspective, changes in liability are as significant as changes in accident risk. Generally, autonomous driving technology is expected to reduce accident frequency. Research analyzing insurance claims data from Waymo vehicles indicates that the number of personal injury and property damage claims is significantly lower than that of conventional vehicles. However, a decrease in accident frequency does not necessarily mean a reduction in insurance risk. In fact, the cost of damages per accident may increase. The advanced equipment in autonomous vehicles, such as lidar, radar, and cameras, can raise repair costs, and the costs associated with data analysis for determining accident causes and legal disputes may also rise. Thus, while accident frequency may decline, the risk structure is likely to shift toward higher costs per incident.

In this context, the importance of vehicle data becomes even more pronounced. To determine the causes of accidents involving autonomous vehicles, access to driving records and system operation information is essential. If data access is restricted, it can complicate liability determinations and increase investigation and litigation costs. Therefore, standardizing accident-related data and establishing a framework for reasonable use by stakeholders, including insurance companies, is crucial.

The success of autonomous vehicle commercialization will not be determined solely by technological advancements. A societal consensus on who is responsible when accidents occur and how victims will be protected must also be reached. This requires establishing a liability framework for each stage of autonomous driving, legislating data access rights, and standardizing records. Insurance in the era of autonomous vehicles should serve as a safety mechanism that provides societal trust in new technologies, going beyond mere accident compensation. As we approach the commercialization of autonomous vehicles, it is imperative to engage in more specific discussions regarding liability and compensation frameworks.

* This article has been translated by AI.