On August 27, 2026, Chinese media CNMO Tech reported that at the "AI Self-Driving Mobility Demonstration City Success Strategy Forum" held in Gwangju, South Korea, a series of points were made that the development of the autonomous driving industry requires not only technological development but also the simultaneous advancement of building profitable business models and integrating with existing transportation and logistics services.

The forum was held against the backdrop of a large-scale demonstration project that designates the entire city of Gwangju as an autonomous driving demonstration zone and utilizes 200 Level 4 autonomous vehicles based on Hyundai Motor's "IONIQ 5." More than 100 experts and stakeholders participated.

Han Ji-hyun, CEO of Autonomous A2Z, a startup founded by engineers from Hyundai Motor's research institute, pointed out that companies ultimately need to generate profits and cannot rely on subsidies for research and development long-term. He argued that instead of waiting for the technology to fully mature, the existing transportation industry, including taxis, public transportation, and logistics, should be involved in commercialization from the demonstration stage, and integration with services should be promoted. He also presented the outlook that it will still take 3 to 4 years to provide practical autonomous driving services.

Lee Dong-hyun, Director of Gwangju City's Future Mobility Industry Bureau, explained that the purpose of the demonstration project is not only to verify the operational capabilities of vehicles but also to accumulate driving data necessary for end-to-end (E2E) autonomous driving. In the E2E method, vast amounts of driving data are used to train AI, allowing it to directly generate driving decisions. Gwangju City plans to expand the demonstration scale in the future and aim for fully unmanned operation while gradually reducing the number of drivers.

Professor Lee Ki-bum of Gachon University pointed out that future autonomous driving competition will not only be about algorithm performance but also a "data competition" to collect and accumulate high-quality data and use it for effective learning. He emphasized that data from "edge scenarios" such as emergencies that could lead to accidents, complex intersections, and non-standard roads is particularly important. If such data is shared among vehicles, the unique situations experienced by one vehicle can be used for the learning of the entire system.

Experts also raised the opinion that large-scale AI infrastructure is needed to utilize vast amounts of driving data for AI learning. Furthermore, in preparation for accidents occurring on public roads, establishing a system in advance to analyze operational data and clarify the cause and responsibility was pointed out as an important challenge for the full-scale commercialization of autonomous driving in the future.