Every time the government announces it will invest trillions of won in the AI industry, one phrase is never missing from the press release: "job creation." But is it true? To answer that, we first have to dust off an old number.

It's called the employment-inducement coefficient.

It measures the number of wage earners created—directly and indirectly, across an industry and its related sectors—when one billion won of final demand arises in that industry. Calculated by the Bank of Korea on the basis of its input-output tables, the coefficient is the most intuitive yardstick we have for how many people's livelihoods a given industry actually sustains.

As of 2023, the coefficient for Korean industry as a whole stood at 6.2. In other words, every billion won of new demand pulls roughly 6.2 jobs along with it.

The trouble is that this number varies wildly from one industry to the next.

Semiconductors yield 2 jobs per billion won; education yields 13

Between services and manufacturing lies a gulf that cannot be crossed. Labor-intensive services—education, health and social care, food and lodging—post coefficients well above 10. Construction does too. In industries that require human hands, money turns into jobs almost as soon as it arrives.

Advanced manufacturing, by contrast, tells a different story. Sectors like semiconductors and electronic components are capital-intensive and heavily automated, so even a billion won of investment yields only two or three jobs. Software and IT services, too, absorb far less labor than traditional services do—because their whole structure is built on a small number of people generating high productivity.

The AI industry pushes this tendency to its extreme. A handful of GPU clusters and a small team of engineers can produce hundreds of billions of won in revenue. The headcount-to-revenue ratio at an AI firm barely compares with that of a traditional industry. According to a December 2025 release from the Bank of Korea, Korea's pool of dedicated AI specialists was estimated at roughly 57,000 people as of 2024—a startlingly small figure for an industry attracting trillions of won in investment.

It's already happening

Let's set theory aside and look at reality. A report published in October 2025 by the Bank of Korea's employment research team, "The Spread of AI and the Contraction of Youth Employment," proves the point in numbers.

In the three years after ChatGPT's launch (July 2022 to July 2025), 211,000 jobs held by young people (ages 15–29) vanished. Of those, 208,000 disappeared in industries highly exposed to AI—98.6% of the total decline.

The picture sharpens when you zoom in. Youth employment in computer programming and systems integration, which includes software development, fell by 11.2%. Publishing dropped 20.4%, information services 23.8%. Even professional services like law and accounting shed 8.8%.

What's striking is the movement among workers in their fifties. Over the same period, jobs held by people in their fifties grew by 209,000, and 146,000 of those were in AI-exposed industries. The Bank of Korea called this "seniority-biased technological change." Its analysis: as AI replaces the routine work of junior employees, the organizational and judgment skills of senior workers have, if anything, become scarcer and more valuable.

To sum up: the AI industry creates few jobs of its own. At the same time, in the other industries where AI spreads, it shrinks the jobs that already exist. And it pushes out young people trying to enter the labor market first.

Measuring a new era with an old ruler

And yet the government's industrial policy still leans on the old formula: investment leads to growth leads to jobs. Trillions of won for AI semiconductors, trillions for AI data centers, hundreds of billions to nurture AI startups. These investments will surely do something to lift GDP. But the employment-inducement coefficient is far too low to justify hanging a "job creation" sign over them.

The coefficient itself has its limits, too. Because it is built on the input-output tables, it fails to capture the ripple effects of an AI that spills across the boundaries of industrial classification. AI raises manufacturing productivity and cuts jobs in one place while spawning entirely new occupations in another—and that process simply cannot be measured with the existing coefficient.

Which is to say: if we want to measure the employment outcomes of a nation's industrial investment, the employment-inducement coefficient alone is no longer enough.

We need a new yardstick

Employment policy in the age of AI needs at least three supplementary metrics.

First, a "job-conversion coefficient." We need to measure how much investment in one industry reduces or transforms employment in others. If a billion won invested in AI creates three jobs within the AI sector while wiping out five in traditional services, the net employment effect is negative. We have to abandon the habit of calculating an investment's employment impact solely within the industry receiving it.

Second, a "job-quality index." We need to look at the wage level, job security, and skill requirements of the jobs being created. If AI creates few jobs but offers high pay and stability, that warrants an assessment different from a simple head count. Conversely, if the jobs AI displaces are merely being converted into platform or gig-economy work, that is not job "creation" but job "degradation."

Third, a "generational employment-impact assessment." As the Bank of Korea's research showed, AI's employment shock breaks in opposite directions by generation. For workers in their fifties it becomes an opportunity; for those in their twenties, a barrier. If investment policy ends up blocking a particular generation from entering the labor market, that is not merely a failure of industrial policy—it is a social hazard.

Change the criteria for investment, not its direction

This is not an argument against investing in AI. Falling behind in the AI race would put the future of the entire economy at risk. That much is clear.

But the practice of measuring the returns on AI investment by "how many jobs it created" is now an anachronism. Slapping the label "job creation" on an industry whose coefficient is just two or three is dishonest—to the public, and to the policy itself.

It is time, instead, to redesign the performance metrics for AI investment policy honestly. We should put indicators like contribution to productivity gains, industrial-competitiveness indices, and technological self-sufficiency front and center, and separate the question of employment onto its own policy axis. A more realistic approach would be to design a structure that redistributes part of the value AI creates into fields with high employment-inducement coefficients—education, care work, health.

How many people will be put to work when you pour in a billion won? If this old question still holds in the age of AI, then we have to change the way we answer it. You cannot measure a new era with an old ruler.