The Bank of Korea said on the 18th that an analysis found that of the 285,000 youth jobs (ages 15 to 29) lost over the four years since ChatGPT was released in 2022, 94% (268,000) were jobs exposed to artificial intelligence (AI), such as computer programming. As core tasks became replaceable by AI, corporations reduced new hiring.
According to the report released the same day by the Bank of Korea, "Is AI to blame for youth employment contraction? The changing career ladder and policy tasks," as of June, youth jobs in information services fell 31.4% compared with June 2022, four years earlier. Publishing fell 27.4%, computer programming fell 16.6%, and professional services fell 11.6%. All are jobs in which AI can replace some or all tasks.
By contrast, jobs for people in their 50s increased by 230,000 on the same basis, and 75.2% (173,000) of those were in AI-exposed industries. Jobs for seniors with extensive careers and experience are increasing, while youth jobs assigned to simple tasks that can be replaced by AI at any time are decreasing, showing a "seniority bias."
In particular, the unemployment rate of highly educated youth in high AI-exposure fields is also on the rise. According to the Bank of Korea, since November 2022, when ChatGPT was commercialized, the average unemployment rate for college-educated youth was 7%, 1.6 percentage points higher than the 5.4% for youth with a junior college education or less. This differs from the past, when unemployment rates for college-educated youth and those with a junior college education or less were similar. In contrast, among middle-aged and older adults, unemployment rates by education level showed little change before and after the release of ChatGPT.
The number of young people leaving (outflows) into unemployment from high AI-exposure industries has also increased, the analysis found. When the Bank of Korea grouped all industries into four categories by degree of AI exposure, outflows in the two high-exposure categories rose 32.4%, from 3,700 in the four years from January 2016 to December 2019 to 4,900 in the four years from July 2022 to June 2026. In contrast, the two low-exposure categories rose only 6%, from 6,600 to 7,000.
Some note that not all causes of worsening youth employment are due to AI. They say it is the result of a stronger preference for experienced hires over spending expense to train new hires until they can take on key tasks. On this point, a Bank of Korea official said, "There was a preference for hiring experienced workers, and with the advent of the technology called AI, the most reasonable inference is that it accelerated this."