SEOUL, March 27 (AJP) – Jenny Kim, 23, a senior at Yonsei University, had thought finding a job would be secure once she entered a top Seoul university majoring in engineering, but many of her peers are now despondent.
“There may not be much left to do in computer engineering,” she said, as artificial intelligence has already replaced much of the entry-level back-office corporate work.
Sophomore Kim S.H., studying structural engineering, also said he and his peers are increasingly exploring career paths unrelated to their majors as the field is rapidly being reshaped by AI.
College graduates in Korea now vie against AI in an already tight job market, with the average time to land a first job stretching to nearly nine months — the longest in 20 years, according to a recent survey as of May 2025. Even degrees from top universities or STEM majors no longer guarantee employment.
“Many returning from military service delay graduation as long as possible to buy time to apply for jobs,” said J.W. Park, 29, a Yonsei graduate who landed a job 18 months after submitting nearly 100 applications.
The employment ratio of Yonsei University graduates fell to 46.6 percent last year from 51 percent in 2024.
AI is now cutting into hiring across South Korea’s traditionally secure professions, with entry-level roles shrinking in law, accounting and technology — forcing young graduates to compete not only with each other, but increasingly with machines.
The strain is already visible in the labor market. Employment among those aged 25 to 29 fell to 2.346 million in February, the lowest for the month since 2017, while the employment rate slipped to 70.4 percent. Youth unemployment rose to 7.1 percent, with underemployment reaching 17.4 percent, the highest February level in three years.
Losses have been concentrated in white-collar sectors. Employment in information and communications fell by 52,000 — the steepest drop since 2014 — while professional, scientific and technical services, including legal and accounting, declined by 29,000.
Studies point to AI as a key driver. The Korea Employment Information Service found white-collar jobs more exposed to AI-driven displacement, identifying lawyers, accountants and journalists among the most at risk.
A Bank of Korea report showed that of 211,000 youth jobs lost between July 2022 and July 2025, about 208,000 were in highly AI-exposed sectors such as programming.
“AI can relatively easily replace routine tasks typically performed by less-experienced workers,” the central bank said.
That shift is most visible at the bottom of the hiring ladder.
“We received around 120 applications for just two positions,” said Rhee Jay Jun, managing director at Young & Jin Tax Consulting Corp.
“Whether they stay depends on how well they can filter out AI errors,” he added, noting that junior roles are shifting from execution to verification.
The pressure is compounded by a bottleneck in certification. Of 12,263 candidates who sat for this year’s CPA preliminary exam, only around 1,200 are expected to pass, and even fewer will secure the mandatory training positions required to qualify fully.
Yet even as AI displaces routine work, its limits remain clear.
In one tax appeal involving hundreds of billions of won, an AI system cited a non-existent Supreme Court ruling — a so-called “hallucination” — nearly leading to a flawed argument.
“We have to fact-check everything again,” Rhee said.
A similar tension is playing out in the legal sector, where AI is lowering barriers while intensifying competition.
Nearly 90 percent of civil first-instance cases in 2024 involved at least one self-represented party, according to the Supreme Court of Korea, with generative AI increasingly used to draft complaints and briefs.
At the same time, the number of registered lawyers has surged to 38,123, up 76 percent from 2016.
“Giving the same materials to a junior associate and to Gemini, Gemini often does a better job,” said Kim Woong, managing partner at Namdang Law Firm.
But within the legal profession, views diverge on how far AI will reshape hiring.
Ko Eunyoung, a lawyer at Barun Law, said AI cannot replicate “practical experience” and “judgment,” stressing that strategic decision-making in complex cases remains firmly within the domain of human lawyers.
She noted that large firms are better positioned to continue hiring. “Hiring a few junior lawyers is less financially burdensome for bigger firms,” she said, allowing them to maintain recruitment even as smaller firms scale back.
Ko emphasized that legal work requires judgment at every stage of a case. While AI can assist with document review and analyzing opposing arguments, overall case management still depends on human experience, strategy and real-time decision-making.
She also warned of longer-term risks. Reduced hiring at the junior level could weaken the profession’s pipeline, leaving fewer lawyers able to develop into mid-level practitioners.
Academic experts share a similar view.
“Because law is as much about persuasion and judgment as it is about raw information review and analysis, I am confident that it will survive this technological change,” said Brendan Ballou-Kelley of Stanford Law School, a former U.S. federal prosecutor.
He cautioned, however, that AI remains less reliable in drafting briefs and should not replace judicial reasoning.
Mark A. Lemley of Stanford Law School said AI adoption in U.S. legal practice has already expanded rapidly, with litigators using it to draft briefs and corporate lawyers to prepare contracts.
But he warned of growing risks tied to AI errors. “We have seen over 800 cases in which lawyers have been caught filing briefs that use hallucinated citations,” he said.
Experts say the broader question is no longer whether AI will reshape jobs — but how far it will go.
Some warn that rapid AI adoption could deepen inequality and trigger economic disruption, as gains concentrate among those who control capital and technology, while others point to risks of overinvestment and asset bubbles.
For young professionals in Korea, however, the disruption is already here and for jobseekers, the real task would be not just beating other candidates, but avoiding areas machines can do better.
Business-major 21-year-old Gonhee says she would settle for any big-company job that won't be affected by AI.
Copyright ⓒ Aju Press All rights reserved.