ACKNOWLEDGEMENTS

This white paper is the result of a symposium on AI and the future of labor, convened by the Keough School of Global Affairs at the University of Notre Dame and the Seoul National University AI Policy Initiative (SAPI), and held at Seoul National University’s School of Law on February 4, 2026. The authors* gratefully acknowledge the support of the Korea Information Society Development Institute (KISDI), the SNU Center for Trustworthy AI, and the Korean Association for Artificial Intelligence and Law (KAAIL). The symposium was supported by the Korea Foundation.The views expressed in this white paper are those of the authors and do not necessarily reflect those of the University of Notre Dame, Seoul National University, KISDI, or any affiliated institutions.

The symposium was conducted under the Chatham House Rule. Accordingly, while this paper draws on the substance of the discussions, specific remarks are not attributed to individual participants. The following is a non-exhaustive list of the symposium participants who contributed to the dialogue and findings reflected in this report. Some participants chose to remain anonymous.

AUM Sangmin  Kyung Hee University

BAN Ga-Woon  Korea Research Institute for Vocational Education and Training

CHOI Sukhwan  Seoul National University

HAN Joseph  Korea Development Institute

KANG Namkyu  Gaon Law Group

KIM Eunsoo  Seoul National University

KIM Hyangmi  LG AI Research

KIM Hyounju  Korean Confederation of Trade Unions

KIM Hyunjung  IBM Korea Consulting

KIM Suji  Korean Confederation of Trade Unions

KIM Yeweon  Seoul National University

KO Haksoo  Seoul National University

LEE Hwanoong  Konkuk University

LEE Yong Suk  Keough School of Global Affairs, University of Notre Dame

LIM Yong  School of Law, Seoul National University

MOON Ahram  AI Economic Policy Group, Korea Information Society Development Institute

PARK Joonwoo  Seoul National University

PARK Woochul  NAVER

SHIN Yongseok  Washington University in St. Louis

SUH Donghyun  Bank of Korea

YOU Jaeyoun  Hanyang University

YU Sumin  Seoul National University

YOON Chan  Microsoft Korea

YOUN Hyejin  Seoul National University

_________________________________________________________________________

EXECUTIVE SUMMARY

The rapid advancement of artificial intelligence technologies presents both significant opportunities and profound risks for workers and economies around the world. South Korea, as one of the world’s most technologically advanced and economically dynamic nations, faces a distinctive set of challenges at the intersection of AI and labor. This white paper examines these challenges through the lens of the Korean context, drawing on an expert symposium held at Seoul National University on February 4, 2026, that brought together academics, industry leaders, labor representatives, government officials, and legal professionals.

AI’s labor market effects are increasingly being felt in Korea – areas identified through the symposium included financial services call centers, the legal profession, and software development. Population-level employment data reveals that a significant increase in an occupation’s exposure to predictive AI was associated with roughly a 4.6% reduction in employment, concentrated among middle-aged male workers in manufacturing. Generative AI’s effects, while not yet statistically significant in aggregate data through 2024, are clearly emerging at the firm level: some 63.5% of Korean workers have used generative AI, with average time savings of approximately 3.8% of working hours among users. However, freed up time goes to both productive work and workplace leisure time. Access to AI tools and their productivity benefits is also stratified by income, firm size, and employment type.

The upskilling paradox is a central labor challenge. The initial expectation was that AI would primarily threaten less tech savvy and soon-to-be redundant managing level workers. The emerging evidence tells a less anticipated side to the story: one of the hardest-hit groups is young, entry-level workers who have not yet had the opportunity to develop the requisite skills and deep professional expertise that would enable them to advance and thrive in the workplace. As AI enables firms to achieve adequate performance with fewer junior workers, the traditional skill ladder—the progression from novice to expert through on-the-job learning—is eroding. The legal sector exemplifies this paradox: AI allows more seasoned lawyers to work more efficiently, e.g., complete arduous tasks such as research and drafting more efficiently with less help from junior attorneys. This, however, can reduce invaluable learning and training opportunities for junior attorneys that would have helped them develop expertise needed for senior roles. In call centers, experienced agents can identify AI errors that newer, post-AI-trained agents cannot. The upskilling problem—preserving the institutions through which workers develop deep expertise—is as urgent as the reskilling problem of teaching displaced workers entirely new skills.

Korea’s dual labor market structure amplifies AI-related risks. The rigid divide between protected regular workers and vulnerable non-regular and subcontracted workers means that the burdens of AI-driven displacement can fall disproportionately on those with the least bargaining power. Call center workers (customer service representatives) have already lost jobs. Other sectors and workers may soon follow.

Korea’s macro-economic context makes the AI transition especially challenging. The country’s growth model faces simultaneous pressures: intensifying competition from China, the world’s lowest fertility rate, K-shaped growth that feeds a zero-sum social mentality, and institutional rigidity rooted in the legacy of government-led development. The Nokia cautionary tale applies: the company went from 50% global market share to collapse in two years because its commitment to the existing model suppressed voices of warning. Korea’s economic establishment may face an analogous risk if it doesn’t adapt.

Workshop participants broadly agreed that Korea is heading toward a “technological caste system.” In the scenario exercise, all three breakout groups identified augmented AI capability paired with passive societal response, producing deepening inequality, as the most likely outcome under current conditions. This is not an “ability gap” but a “structural exclusion” problem. The decisive dimension is not how AI develops but whether society responds actively or passively.

The speed of AI development far outpaces institutional adaptation. Korea’s AI policy has focused on promoting technological innovation and economic growth, while concrete labor market policies remain underdeveloped. The Economic, Social and Labor Council has only reached the “question-framing” stage on AI. Major labor unions have disengaged from formal dialogue processes, and youth voices are absent.

The paper proposes policy options across seven domains:

• institutionalizing multi-stakeholder dialogue on AI and labor, with youth representation and actionable outcomes;
• boosting government communication and public awareness of the potential impact of AI on labor—recognizing that communication itself is policy;
• promoting research, data, and measurement that sheds light on the real time effects—including labor market outcome and productivity metrics;
• undertaking social safety net reform—extending unemployment insurance, pursuing job transition flexibility, and examining the data dividend concept;
• investing in workforce development—shifting from formal, compliance-oriented training to genuine capability building;
• pursuing education reform—redesigning higher education to address the looming upskilling deficit, and investing in the meaning of work as an educational theme from childhood; and
• balancing public finance considerations—distinguishing between employment insurance (occupation-specific) and education budgets (foundational), and exploring data revenues as a supplementary source.

A cross-cutting theme was the need for a new social compact on workforce development. The existing system in which the state provides basic education, universities produce cultivated citizens, and firms hire and train on the job is breaking down. A new agreement among workers, industry, educational institutions, and government is needed, built through multi-stakeholder dialogue conducted before the issue becomes politically polarized.

Korea has the intellectual resources, institutional capacity, and historical experience to navigate this transition. What is needed is the political will to act and the honesty to recognize that past models, however successful, may not serve the future.