“AI 너무 빨라” 속도조절 나선 빅테크…‘안전’ 뒤 숨은 셈법은

South Korea's AI proliferation is causing a more pronounced concentration of jobs in the capital region, a new analysis reveals. The Korean Bank's regional economic research team released a report titled "AI and Regional Labor Markets" on May 15, categorizing artificial intelligence into three types: "generative," "agency," and "physical." The report, presented at the 2026 BOK Regional Economy Symposium in Daejeon, provides insights into AI exposure by occupation and region.

Generative AI exposure was highest in Seoul, Sejong, Gyeonggi, and Incheon, while agency AI exposure was highest in Seoul, Sejong, Gyeonggi, and Daejeon. Both categories showed high exposure in the capital region. The exposure rate represents potential automation or the level of AI utilization, calculated based on data from 2024's business activity survey and the 2025 economic activity population survey.

Physical AI exposure, which involves sensors perceiving the real world and actuators (motors, etc.) executing physical actions, was highest in Gyeongsangbuk-do, Jeollanam-do, Chungbuk, and Ulsan, south of the capital region.

The report indicates that high AI exposure in the capital region correlates with a greater concentration of jobs in writing, sales, management, and specialized professions. In contrast, jobs with lower AI exposure in non-metropolitan areas included machinery operation, simple labor, and service jobs. This disparity in job growth is expected to widen the gap between the capital and other regions.

From 2023 to 2025, as generative AI began to spread, employment for those exposed to generative AI increased by 247,000 (ages 33-59), with the capital region accounting for 80.8% (19,900,000). Agency AI exposure also saw a rise in job opportunities in 2025, with the capital region holding an 88.3% share (9,300,000) of the increase. The labor impact of physical AI exposure has been declining in both regions for a long time.

The team attributes the regional job growth to disparities in AI utilization and information technology human capital. They also note that employment for 20-year-olds, who are entering the workforce without specialized skills or experience, is not significantly different between the capital and non-capital regions. The researchers recommend strengthening AI education systems through collaboration between universities and public institutions, promoting the fusion of manufacturing and knowledge services, and supporting the establishment of start-ups.

They suggest that regional university hubs could play a central role in AI research and education, urging increased investment in this direction.