Generative artificial intelligence is more likely to first squeeze entry-level jobs for highly educated professionals such as accountants and lawyers, according to an analysis by the Bank of Korea. Rather than replacing professionals outright, AI is first taking over the data organization, document drafting and basic analysis that newcomers have long handled, which could reduce the chances for new hires to gain experience.
Youth jobs for those aged 15 to 34 fell by 285,000 between June 2022 and this June, according to a report titled "Is AI to Blame for the Slump in Youth Employment?" released by the employment research team of the Bank of Korea's Economic Research Institute. Of that decline, 268,000 came in industries highly exposed to AI, accounting for 94% of the total drop. Over the same period, jobs for workers in their 50s rose by 230,000, of which 173,000 were added in industries highly exposed to AI.
The analysis found that a seniority-biased shift — in which new and junior workers decline while experienced workers increase in industries that make heavy use of AI — has continued in the latest data. Youth employment in industries highly exposed to AI has fallen sharply. Compared with June 2022, professional services fell 11.6%, information services dropped 31.4%, publishing declined 27.4%, and computer programming, system integration and management fell 16.6%.
Employment conditions also worsened relatively for highly educated young people, who face relatively higher exposure to AI. Since November 2022, the average unemployment rate for young college graduates stood at 7.0%, 1.6 percentage points higher than the 5.4% for young people with a junior-college education or less. From January 2019 to October 2022, before ChatGPT's release, the average unemployment rates for young college graduates and those with a junior-college education or less were 8.2% and 8.0%, respectively, showing little difference. More recently, however, employment for young college graduates has taken a bigger hit.
Indeed, AI exposure among workers with five years of experience or less rises with education level: 0.23 for high-school graduates, 0.25 for junior-college graduates, and 0.27 for those with a four-year college degree or higher.
Differences also emerged depending on how AI was adopted. Industries with a higher share of automation — where AI directly replaces human work — saw larger declines in youth employment, while this pattern was less pronounced in industries with a higher share of augmentation, where people use AI to raise productivity. The monthly average outflow of young workers from industries highly exposed to AI rose about 32%, from 3,700 in 2016-2019 to 4,900 between July 2022 and June 2026, while new inflows fell from 32,600 to 29,100.
This suggests that generative AI, by first automating the tasks that newcomers have typically handled — such as searching and organizing data, drafting documents, basic analysis, translation and simple coding — could weaken the first rung of the career ladder.
Still, the Bank of Korea stressed that it is difficult to attribute the recent weakness in youth employment to AI alone. "Other factors, such as the spread of remote work since the pandemic, may also have played a role," the bank said. "It is more appropriate to view the current slump in youth employment not as a direct effect of AI but as AI accelerating changes in existing hiring practices."
