In this blog post, we’ll examine the challenges that autonomous driving technology must address in the future—from technological advancements and ethical issues to the potential of connected cars, solutions to traffic congestion, and changes in social systems.
Recent technological advancements—driven by the convergence of big data, the Internet of Things (IoT), and artificial intelligence—have reached a level where devices can independently assess their surroundings and perform tasks. Among these, interest in self-driving cars—vehicles that operate autonomously without human intervention—is exploding worldwide. Just as smartphones have drastically transformed the world in a short period, there is great anticipation that smart cars—specifically, self-driving cars—will also bring numerous benefits to our lives. Global automotive giants like Tesla, along with Waymo—Google’s self-driving technology subsidiary—major IT companies such as Apple, and network-related firms like SK Telecom have all been involved in the development of autonomous driving technologies and related businesses to secure a competitive edge in the new technology and mobility markets. Autonomous vehicles, which can be described as a convergence of cutting-edge technologies such as artificial intelligence, computers, communications, and sensors, are a complex system that requires the simultaneous advancement of multiple engineering disciplines to be realized. However, it has also been argued that the obstacles hindering the advancement of autonomous driving technology may not simply be issues of technical development, but rather matters of ethics and social consensus. In particular, questions such as whether to protect the vehicle’s occupants or multiple pedestrians in situations where an accident is unavoidable have been consistently debated as representative ethical dilemmas surrounding autonomous vehicles. The following is a common example often cited when discussing the ethical issues of autonomous driving: It is an extreme scenario in which, when several people suddenly appear in front of an autonomous vehicle, the system must choose between colliding with them or swerving to avoid the accident—thereby putting the occupants at risk. If the autonomous vehicle’s program is set by default to a “self-sacrifice mode” that prioritizes the safety of others over the occupants, many people might hesitate to purchase such a vehicle for their own safety. Conversely, if a self-driving car is designed to prioritize avoiding a collision with others in order to protect its occupants, there is a possibility that social backlash could arise during its launch. However, when encountering such examples, there is one point worth considering: Are these ethical issues truly new problems in moral philosophy that arise only with the introduction of self-driving cars? In everyday life, it is not uncommon for people to appear on the road at unexpected moments. Sometimes pedestrians jaywalk; other times, they’re still crossing the crosswalk after the pedestrian signal has turned red; and sometimes they enter the road in ways that are difficult for drivers to anticipate. If an accident occurs in such situations, liability must be determined by comprehensively evaluating various factors—including the pedestrian’s actions, the driver’s duty of care, and the road conditions—and it cannot be assumed that the pedestrian is always at fault. Even with the introduction of autonomous vehicles, liability for accidents cannot be uniformly attributed to either the vehicle or the pedestrian. It is necessary to comprehensively examine the circumstances at the time of the accident, whether the system was operational, the obligations of the driver or passengers, and the responsibilities of the vehicle manufacturer and the autonomous driving system developer. Therefore, when designing autonomous vehicles, it is important to design systems that prevent accidents as much as possible and minimize damage, rather than simply choosing between a “self-protection mode” or a “protection-of-others mode.”
Some people harbor doubts about the technology behind autonomous vehicles that react to people or objects to avoid accidents. They question whether artificial intelligence can truly replicate the ability of humans to make split-second judgments and choose the best course of action for each situation. However, autonomous driving technology has already moved beyond the simple research stage and has advanced to the point where it is being applied on actual roads and in commercial services. The U.S. National Highway Traffic Safety Administration (NHTSA) currently classifies autonomous driving systems within the advanced automation categories of SAE Levels 3–5. At the same time, the agency explains that even the highest level of driver-assistance features available in vehicles currently on the market still requires the driver’s constant attention and supervision. On the other hand, Level 4 services—where vehicles operate autonomously without human intervention within limited areas and under specific conditions—have already seen real-world commercial applications. For example, Waymo operates a fully autonomous ride-hailing service carrying passengers without a driver in several U.S. cities, and in 2026, it expanded its service to Denver, Tampa, and San Diego. Therefore, it is difficult to view the technology that enables autonomous vehicles to perceive and assess situations as merely a distant possibility. However, technology capable of driving on all roads and in all situations without human intervention has not yet been commercialized, and the reliability and safety of the technology must continue to be verified. Tesla’s current FSD (Supervised) system can also perform various driving functions—such as lane changes, navigating intersections, and turning—but it requires active driver supervision and does not constitute fully autonomous driving. In the autonomous driving market, where massive amounts of manpower and capital are being invested worldwide, the precision of a vehicle’s ability to perceive and assess its surroundings will continue to improve. If this technology develops sufficiently, it has the potential to detect and respond to dangers much faster than humans can; and the technology expected to further enhance this reaction speed and information-sharing capability is the “Connected Car.” The term “Connected Car” refers to technology that enables a vehicle to connect to the internet and communication networks, allowing it to exchange information with other vehicles, traffic infrastructure, and various services. When autonomous vehicles and connected cars are combined, individual vehicles can do more than simply detect and react to their immediate surroundings; they can share information obtained from other vehicles and road infrastructure to gain a broader understanding of traffic conditions. This increases the likelihood of detecting and responding to sudden accidents or road hazards earlier. However, it may take considerable time to reach the ultimate form of autonomous driving technology, where vehicles and networks are tightly integrated. Even currently commercialized driving automation technologies cannot perfectly handle accidents in every situation, and factors such as technological limitations, safety verification, and legal and regulatory standards may influence the pace of autonomous driving adoption. In particular, since a significant portion of the autonomous driving technologies available to consumers today require continuous driver supervision, they should not be equated with fully autonomous driving without a driver.
While the expected benefits of autonomous driving technology include driver comfort and freedom of movement, it also holds the potential to improve traffic flow by more efficiently maintaining distances between vehicles and leveraging reaction times that are faster than those of humans. If vehicles can share information such as their location, speed, and direction of travel and move in a coordinated manner, unnecessary sudden braking or acceleration can be reduced, resulting in smoother traffic flow. In particular, major Asian cities, including Seoul, have long struggled with traffic congestion, and in some cases, expanding various modes of transportation—such as bicycles and trams—alone is insufficient to solve all traffic problems. In this context, autonomous vehicles are gaining attention as a potential solution to improve traffic flow. From the perspective of both the national and municipal governments, this technology is difficult to dismiss, as it has the potential to reduce social costs—such as the waste of time and fuel caused by traffic congestion, as well as the risk of accidents. Therefore, we must also consider how to utilize this technology, which is not yet fully developed. One approach could be to first apply autonomous driving systems in relatively predictable environments within city centers—such as vehicle-only lanes—to reduce the complexity the technology must handle while simultaneously addressing traffic congestion. Even in large cities like Seoul, traffic often becomes concentrated on specific roads and major arterial routes, leading to congestion in many areas; however, advancements in vehicle-to-vehicle communication and efficient traffic control technologies could help alleviate these problems. Autonomous vehicles are highly likely to become an increasingly integral part of our daily lives in the future, and if sufficient social consensus is reached regarding the technology’s safety and ethical issues, the scope of autonomous driving services could expand even further. In fact, the fact that driverless ride-hailing services are already commercially available in limited areas demonstrates that autonomous driving is not merely a futuristic fantasy. As of 2026, Waymo is providing fully autonomous ride-hailing services to the general public in several U.S. cities and continues to expand its service areas. However, even if the ethical issues surrounding autonomous driving technology are resolved and actual services begin, that does not mean everything is settled. Since autonomous driving technology is a combination of various engineering disciplines, there are many details that require careful preparation—not only the technology itself but also laws and regulations, road environments, traffic signal systems, insurance and liability standards, and pedestrian safety. These issues must be thoroughly discussed by experts and members of society before autonomous vehicles are fully introduced. For example, as mentioned earlier, a dangerous situation can arise for pedestrians who remain on a crosswalk even after the pedestrian signal has turned red. In particular, since the walking speed and ability to respond to surrounding situations of the elderly, people with disabilities, and children may differ from those of typical adults, they should not be excluded from potentially dangerous situations simply because the signal has changed. Therefore, just as measures such as “pedestrian signal extension buttons” or adjusted crossing times for vulnerable road users have been implemented, it is essential to simultaneously refine the broader social infrastructure—including roads and traffic signal systems—in tandem with the introduction of autonomous driving systems. To ensure that autonomous vehicles do not make mistakes, it is not enough to simply change the vehicles themselves. The roads on which they travel, traffic signals, laws and regulations, and even the behavior of the people sharing those roads must all change as well. Ultimately, the true realization of autonomous driving lies not in a single vehicle making all decisions perfectly, but in creating a transportation system where vehicles, roads, cities, and people are safely interconnected.
