Suddenly, the hottest question in artificial intelligence is no longer whether AI will take our jobs. It is whether, one day, it could kill us.
The prospect remains fiercely disputed and may sound like science fiction. Yet researchers inside some of the world's leading AI companies are now openly warning about the possibility of humans losing control of increasingly autonomous systems, with some putting the risk of human extinction within the next decade above 10 percent.
The warnings have become serious enough to draw bipartisan calls for tougher safeguards in Washington and a scramble among technology companies and policymakers to devise "kill switches" and other mechanisms capable of stopping rogue systems.
South Korea ought to be paying attention.
Not because Seoul should abandon its ambition to become an "AI republic," stop building data centers or retreat from the semiconductor boom that has given the country one of its greatest economic opportunities in decades.
Korea cannot afford to sit out a technological transformation that may determine national competitiveness for a generation. But neither can "we will fall behind" become the answer to every question about AI.
Korea has a familiar tendency to turn technological competition into a national speed contest.
Once the race is declared, disagreement becomes obstruction, caution becomes timidity and almost every problem is answered by spending more money and moving faster. That reflex is particularly dangerous with AI because some of the casualties are no longer hypothetical.
Between June 2022 and June 2026, employment among Koreans aged 15 to 29 fell by 285,000. An extraordinary 268,000 of those lost jobs — 94 percent — came from industries highly exposed to AI, according to the Bank of Korea. Youth employment fell 31.4 percent in information services, 27.4 percent in publishing and 16.6 percent in computer programming. Employment among people in their 50s, meanwhile, continued to increase in many of the same industries.
AI is not simply replacing workers. It risks removing the first rung of the career ladder.
Companies can use AI to perform the research, coding, translation, document preparation, customer support and routine analysis once assigned to junior employees.
Senior workers who already possess judgment, networks and institutional knowledge become more productive. Young people who were supposed to acquire those things through entry-level work may never get through the door.
The result could be a cruel paradox. Korea, already struggling with one of the developed world's toughest environments for young people entering stable employment, could become spectacularly successful at producing the chips and computing infrastructure that make AI possible while allowing AI to hollow out the route by which the next generation enters the professional economy.
Compare the urgency of the two policy responses.
When the subject is semiconductor fabs, AI data centers, GPUs and electricity, the government knows how to hit the accelerator. Seoul has unveiled massive semiconductor and AI infrastructure projects, including plans tied to KRW 550 trillion in private investment and data-center capacity totaling 18.4 gigawatts. AI-service data centers have been brought into expanded strategic investment tax incentives, while public financing is being mobilized for GPUs and computing infrastructure.
The government is not entirely ignoring the employment shock. In July it announced a five-year industrial-transition plan that aims to provide AI-related training to more than 1 million people by 2030 and establish an early-warning system for jobs vulnerable to technological change.
But retraining people after their jobs disappear is not the same thing as deciding what kind of labor market an AI economy should have.
Nor does an AI Basic Act built primarily around promoting innovation and establishing general safeguards answer who bears responsibility when AI eliminates junior positions, makes employment decisions, concentrates income among capital owners and highly skilled incumbents, or leaves a generation permanently behind.
Indeed, lawmakers themselves have already pointed to gaps in protections when AI is used in hiring, evaluation, promotion and other workplace decisions.
But the conversation in Korea has barely begun. The country does not need to choose between AI advancement and AI restraint. It needs to stop pretending that acceleration alone constitutes a strategy.
If people developing the world's most advanced AI systems believe it is worth asking whether their creations could eventually threaten humanity, surely a country betting its industrial future on the technology can ask a more immediate question: Who is being sacrificed while Korea races to become an AI republic — and what exactly is the republic being built for if its people are treated merely as the cost of getting there first?
Copyright ⓒ Aju Press All rights reserved.
