When I talk with young people about finding work, one worry keeps coming up. Even for entry-level openings, employers now ask for hands-on experience and a portfolio, and skill with generative AI has become an important qualification as well. Landing a job requires experience, yet the first jobs where that experience is built are steadily disappearing.
The shift I had been sensing on the ground is now beginning to show up in labor-market figures.
According to a recent Bank of Korea report titled "Is AI Behind the Slump in Youth Employment? The Changing Career Ladder and How to Respond," the number of jobs held by people aged 15 to 29 fell by 285,000 between June 2022 and June this year. Of that decline, 268,000 came in industries with a high potential for AI use. Employment also dropped sharply in fields that relatively well-educated young people have entered, such as information services, publishing, programming and professional services.
Of course, not all of this is due to AI. Companies scaling back open recruitment and leaning more toward experienced hires is a long-running trend. But from what I see on the ground, AI is speeding that change considerably. What worries me more is that the jobs disappearing overlap with the ones where young people used to first learn to work.
Young people usually do not take on important decisions from the moment they start working. They handled small tasks, put together materials, and learned on the job as someone told them to go back and fix things. Most of the people I met at work were similar. Their careers were built not because they were skilled from the start, but because they had the chance to try the work.
Now, though, young people face the opposite demand. Getting a job requires experience, yet the entry-level jobs where that experience can be built are shrinking.
That gap becomes clearer in conversations with young people. They earn certifications, take courses, build portfolios, and now must add AI skills on top. What many find hardest is not only figuring out "what to prepare" but also "where I can actually work." More important, the minimum threshold young people must clear just to enter the labor market is itself rising.
The Seoul Metropolitan Government's G3 Seoul planning committee recently discussed a "Youth AI Ladder." In fleshing out the policy, the city said it would provide AI licenses and support AI workspaces and job- and startup-training programs. It is meaningful in that it widens access to new technology. Still, one question kept surfacing during the discussion.
"Is giving someone the chance to use AI the same as giving them the chance to work in the age of AI?"
Even if young people become skilled with generative AI, the jobs problem is not solved unless companies actually hire entry-level workers. What is learned in an AI classroom is an entirely different experience from listening to a customer's needs at a real company, working through friction with colleagues and taking responsibility for one's own output.
Future youth employment policy must not stop at simply increasing the number of AI courses. What matters more is how to expand "real work experience" — carrying out actual projects, working alongside a senior colleague for a set period, and receiving feedback on the results.
One other striking point: over the same period, jobs held by young people fell, but jobs held by people in their 50s rose.
The presidential Economic, Social and Labor Council's special committee on the changing population structure and jobs is also discussing continued employment for older workers alongside young people's entry into the labor market. Here, too, it comes down to the same question: does older workers staying on the job longer reduce jobs for the young?
What is needed is not a way to make the two generations split the jobs between them, but a way to let them work together. For example, a skilled worker takes charge of an actual project while a young worker handles AI-driven research, analysis and execution, producing results together.
There is no stopping AI's advance. Nor can the problem be solved simply by demanding that companies hire more people, because no business can run without weighing costs and productivity. That is exactly why policy has a role to play.
More important than teaching young people one more way to use AI is creating the first chance for them to actually do real work alongside it. We must make sure the process does not vanish — taking on small tasks, failing, revising, learning from someone, and eventually becoming a person with experience.
The biggest problem for youth employment in the age of AI may not be that a few jobs disappear. The greater problem is that the first positions where "someone without experience" could become "someone with experience" are disappearing. Future youth employment policy should be judged not by how many young people were taught AI, but by how many were given a genuine first chance to work and grow in the age of AI.
