21SEP

Korea finds the missing youth jobs

3 min read

16:45UTC

The Bank of Korea counted 285,000 young workers out of insured employment in four years, and found 94% of the loss sitting in sectors it classifies as highly exposed to AI.

← Back to AI closes the hiring door, not the exit Jump to analysis ↓

EconomicDeveloping

Key takeaway

Korea counted young workers leaving insured jobs, and 94% of the loss sat in AI-exposed sectors.

The Bank of Korea published an issue note on 18 August reporting that net youth employment in Korea fell by 285,000 between June 2022 and June 2026, and that 268,000 of that fall, 94% of it, sat in sectors the central bank classifies as highly exposed to artificial intelligence (AI). ¹ Information services fell 31.4%, publishing 27.4%, and computer programming and systems integration 16.6%.

The central bank worked from National Pension Service enrolment records rather than from a survey. It watched people enter and leave insured jobs instead of asking them what they did for a living. Almost every earlier measurement on this beat has either scored how much of a job a model could in principle perform, or counted what employers chose to announce: one public tracker had logged 267 redundancy events and 185,894 affected workers by mid-June, with 56% of them citing AI . A pension register answers a narrower question and answers it precisely, because payroll enrolment is a legal fact rather than a self-description.

Employment among Korean workers in their fifties rose by 230,000 across the same four years, and 173,000 of that gain landed in the same exposed sectors. Young workers vanish where the over-fifties accumulate, inside one set of industries, over one window. Graduate unemployment in Korea has averaged 7.0% since November 2022 against 5.4% among workers holding sub-degree qualifications, which puts the burden on the people the career ladder is meant to admit rather than on those already standing on it.

Two limits travel with the finding. The note's own authors decline to blame AI alone and point to Korea's demographic decline, and a sectoral classification cannot separate substitution by software from the contraction in information services and publishing that hit most rich economies after 2022 for reasons including interest rates and the end of the pandemic hiring bubble. The exact note could not be retrieved from the central bank's own portal, so the figures here rest on outlets that read it and quote its method. What survives both caveats is the shape: a register can say who left insured work and from where, even when it cannot say why.

Deep Analysis

In plain English

The Bank of Korea is Korea's central bank, and the National Pension Service is the country's mandatory state pension scheme: almost every formal job pays into it. Because the bank could count exactly who entered and left insured jobs, rather than asking people in a survey, it found that of the 285,000 net youth jobs lost since 2022, 268,000 sat in industries it classifies as heavily exposed to AI, such as publishing and computer programming. Over the same four years, workers in their fifties in those same industries gained 230,000 jobs. In plain terms: the youngest workers are losing ground in the industries the AI wave has touched, while the oldest workers in those same industries are keeping or gaining it.

Deep Analysis

Root Causes

Korea's National Pension Service records only insured, formal employment, and formal-sector firms are also the ones with the payroll headcount to trim when a sector slows, so the pension data will always show losses concentrated where formal firms sit, whether or not AI is the specific mechanism.

Korea's seniority-based wage system and strong dismissal protection for existing regular employees raise the cost of firing a current worker relative to simply not replacing one who leaves. That structure makes new hiring, not existing headcount, the shock absorber for any slowdown, which is why a fall lands almost entirely on workers in their twenties while the fifties cohort grows in the same sectors.

What could happen next?

Meaning
Administrative payroll data can now separate where AI-linked job losses concentrate from survey-based guesses about who is affected.
Immediate · Assessed
Consequence
If the pattern holds outside Korea, youth unemployment in other administratively-rich economies becomes the leading indicator to watch, not aggregate headline unemployment.
Medium term · Suggested
Risk
A generation excluded from entry-level formal employment in AI-exposed sectors during 2022-26 could carry a permanent earnings gap even after any transition period ends.
Long term · Suggested

Source Landscape

This story draws on neutral-leaning sources

Primary parallel: Japan's 'employment ice age' generation, who left education into the early-1990s recession, was locked out of regular hiring while employers kept existing older staff under lifetime-employment norms. Economists later traced permanently lower lifetime earnings to that one entry cohort, not to workers who already held jobs, the same generational split Korea's pension data now shows.

Counter-parallel: No case in the AI-employment literature has yet shown a youth cohort recovering the entry rungs an automation-linked freeze removed. Every downturn-cohort study to date has confirmed rather than reversed the scarring diagnosis, so there is no positive parallel to set against Korea's numbers.

Consensus view: The Bank of Korea's National Pension Service data confirms what the IMF's January 2026 Staff Discussion Note found using a different method: employment in AI-exposed occupations runs lower once AI adoption spreads through a sector, and the effect concentrates on new entrants rather than incumbents.

Counter-view: Oxford Economics has argued, in research on the broader AI-layoffs trend, that AI's role 'may be overstated', finding little evidence that firms are replacing workers with AI at scale rather than simply freezing hiring during a downturn. Applied to Korea, that reading treats the pension data as a hiring freeze that happens to sit in AI-exposed sectors, not proof that AI itself did the displacing.

Key tension: Whether administrative payroll records that show where jobs disappeared also show why, or whether they capture a correlation between AI exposure and a slowing sector that would have shed junior staff regardless.

Sources:Korea Times

Mentions:South Korea →Japan →Oxford Economics →Bank of Korea →

First Reported In

Update #19 · Four methods, one answer on AI and jobs

Korea Times· 24 Aug 2026

Read original →

Causes and effects

Caused by

New York Fed: adoption up, sackings rare

The New York Fed's hiring-suppression finding locates the mechanism behind Korea's youth job losses rather than refuting them.

Occurred 5 Aug 2026

Read story →

This Event

Korea finds the missing youth jobs

It is the first measurement of AI-era youth job loss taken from an administrative register rather than from an exposure model or an employer announcement.

Led to

BIS, World Bank and ILO agree on AI exposure

Korea's administrative-data finding aligned with three institutions' cross-country measurements published inside the same month.

Occurred 1 Aug 2026

Read story →

Japan's megabanks plan a smaller intake

Bank of Korea's pension-register finding shows the same age-profile pattern from the opposite side.

Occurred 24 Aug 2026

Read story →

Different Perspectives

Salesforce, Synopsys and TD Bank Group

Salesforce, Synopsys and TD Bank Group each filed quarterly disclosures in late August booking restructuring charges, or none at all, without naming AI as a cause. Their silence matters because Challenger's tracker shows AI as a stated reason fell to fourth place in August even as the year-to-date AI-cut total still leads at 116,175.

Singapore, South Korea, Taiwan and Indonesia

Singapore launched its Skills and Workforce Development Agency on 16 September, giving citizens six months of free premium AI tools, while South Korea ring-fenced its AI tax windfall in a new Future Response Fund. Taiwan kept funding its AI build past NT$190bn and Indonesia rewired vocational training around AI literacy, betting state-built skills beat a market-led adjustment.

ver.di, CGT Fonction Publique and CCOO

Germany's ver.di banked a 3.3% pay rise on 1 September and opened talks on a Tarifvertrag Transformation covering dismissal bans and reskilling, while France's CGT rejected Paris's AI negotiating timetable the same week. Spain's CCOO went further on 21 September, proposing to tax companies by the jobs they generate rather than wait for the next bargaining round.

BIS General Manager and Federal Reserve governors

The BIS's General Manager said on 10 September that AI displacement remains limited, even as the BIS's own survey found nearly 80% of firms plan to automate roles. Two Federal Reserve governors made the same point in July, arguing the labour-market data does not yet show a mass-firing event.

Bank of Canada, ONS and ECB

The Bank of Canada found the job-finding gap between AI-exposed and unexposed occupations widened from 2.2 to 13.9 percentage points since 2015-19, while separations barely moved. That framing, a hiring freeze rather than a firing wave, is echoed by the ECB's finding that euro-area AI use hit 52% of workers in 2026, concentrated among the university-educated.

Office for National Statistics

Deferred its Transformed Labour Force Survey beyond November 2027 and disclosed a May 2026 telephone-collection failure. The ONS carries no AI-attribution layer at all, so Britain sits outside this month's cohort of measuring states by its own admission.