Just Because 93% of Jobs Will Still Exist Doesn't Mean You're Safe

The Question Korea's Employment Forecast Never Measured

A qualitative employment outlook published in 2025 by the Korea Employment Information Service (KEIS), a government labor-research institute, contains a figure you might not expect. Projecting employment changes through 2035 across 182 major occupations in Korea, the report forecast a decline for exactly zero of them. More than 93% of the occupations were expected to hold at current levels or grow.

If that number feels jarring, it's because of how the AI-and-jobs conversation has gone for the past decade. It was 2013 when Oxford researchers announced that 47% of US occupations were at risk of automation. More than ten years have passed since, and after generative AI went mainstream in 2022, talk of disappearing professions only intensified. Against that backdrop, a government research institute concluding that zero out of 182 occupations will shrink lands as a genuine surprise.

But before you read this report as grounds for relief, it's worth separating what the figure measures from what it doesn't.

What the Forecast Says: Jobs Aren't Disappearing

The unit of this forecast is headcount — whether the number of people working in a given occupation in 2035 will be higher than, or roughly equal to, today's. By that yardstick, healthcare and caregiving occupations were projected to see steady demand growth as the population ages, and roles in data analysis and AI systems operation also landed in the growth column. Content planning and creative work, along with counseling and coaching, were likewise expected to expand.

The occupations projected to grow share a recognizable pattern: work with stretches that require judgment rather than routine processing, work that demands reading the other person's context, work that involves building trust in ways machines can't easily replicate. By contrast, roles centered on standardized document handling or simple data entry were flagged as far more directly exposed to change. Two pressures — an aging population and accelerating technology — are reshaping the labor market, and the direction of that change is less "humans will do less work" than "the kind of work humans do is shifting."

The survey's scope deserves a note as well: it covers 182 occupations that can be classified as "major occupations" in Korea. Forms of work that resist clean occupational categories — freelancing, solo business ownership, platform-based gig work — simply don't appear in this data. With more than six million solo business owners in Korea and the freelance share of the workforce steadily climbing, you need to understand exactly how much of the labor market this report actually covers before drawing conclusions from it.

What "93% Will Hold" Doesn't Tell You

There is a real counterargument to reading this report as a case for optimism.

A job continuing to exist and the nature of the work an individual does inside that job staying the same are two different stories. Tax accounting may still exist as a profession in 2035 — but that doesn't mean the person working as a tax accountant today will be working the same way, for the same pay. As more people in the market use AI tools to do the same work three times faster, the number of newcomers the field can absorb may shrink. Total headcount holding steady and the barrier to entry rising can happen at the same time. The seats remain, but the density of competition around them changes.

A 2023 OECD report contained an uncomfortable finding on this point: among the occupations most exposed to AI were a large number of high-skill, high-wage jobs. Contrary to the conventional wisdom that simple, repetitive work gets automated first, a substantial share of what professionals actually do — legal review, document drafting, data-driven analysis — has already reached the point where generative AI can handle it. Which means that even if headcounts don't fall, the composition of work inside the professions can change fast.

AI has also kept advancing rapidly since this forecast was published in 2025. A qualitative outlook reflects the technology and industrial structure of the moment it was conducted. By 2027 or 2029, the picture could look quite different — and reading this report's conclusion as a safety guarantee covering the full decade stretches the data well beyond what it can bear.

The fact that a job exists guarantees nothing about an individual's position within it. Seen this way, the 93% figure reads less as reassurance and more as a signal to take stock.

What Solo Founders and One-Person Operators Should Actually Do With This Data

If you want to turn this report into action, the first step is narrowing the question.

Not "will my occupation still exist in 2035" but "will it still exist in the form I currently practice it." Content director may well survive as a job title in 2035. Whether spending 40 hours a week building proposals and reviewing manuscripts survives as a way of doing it is a separate matter. If the market fills with people who produce the same output in 10 hours a week using AI tools, you need to define — before anyone asks — where the other 30 hours of your week create value.

One practical starting point is to actually lay out the list of tasks you currently handle. Go through each item and sort it into two piles: work that AI tools or outsourcing could replace, and work that can't be reproduced at the same quality without your contextual understanding and accumulated relationships. If the second pile is smaller than you expected, your negotiating position can erode even while your occupation survives. The test is simple: pick one task you do now and try producing it at the same quality in half the time using AI tools. If the result holds up, it's time to reassess how much of that task's value was actually yours.

People who have spent decades in one field tend to hit the same question when they audit themselves honestly: "What have I believed I was good at, and does it hold its value without AI tools?" This is exactly the question that gets easier to postpone the longer your career runs and the more stable your job looks. In the accounts of people who prepared for life after retirement, one of the most common regrets is some version of "in all my working years, I never once seriously verified what I was actually good at." A forecast saying jobs won't disappear can, paradoxically, become one more reason to push that reckoning off.

Confirming now which of your capabilities hold up without AI isn't an exercise driven by panic. It's the most practical preparation you can do precisely during a period when your job looks secure.

The KEIS report says that ten years from now, none of Korea's 182 major occupations will employ fewer people than today. Who occupies which of those seats, and on what terms — that's a question the survey never set out to answer. The 93% figure becomes useful only when you treat it not as a verdict of safety, but as the starting point of an audit.