The landscape for South Korean data companies expanding overseas is rapidly changing. While China was once the primary market, Vietnam has emerged as a key player. The surge in demand for labor-intensive data processing due to the expansion of artificial intelligence (AI) has prompted companies to relocate to countries with lower labor costs.
According to the "2025 Data Industry Status Survey Report" released on June 26, 2026, 35.1% of South Korean data companies that generated overseas revenue in 2024 reported earnings from Vietnam. This marks a significant increase of 17.6 percentage points from 17.5% in 2023.
In contrast, the share of revenue from China dropped from 37% to 29.3% during the same period, a decrease of 7.7 percentage points.
Revenue trends mirrored this shift. Earnings from China fell slightly from 104.5 billion won in 2023 to 102.3 billion won in 2024. Meanwhile, revenue from Vietnam surged from 44.5 billion won to 129.2 billion won, nearly tripling and surpassing that of China.
Industry analysts attribute this trend to the characteristics of the data industry, where labor cost competitiveness is paramount. The development of AI services requires extensive data labeling tasks—such as classifying images, videos, audio, and text—most of which are labor-intensive.
The recent shift towards Physical AI has further amplified this trend. Unlike generative AI, which can leverage publicly available data from the web, Physical AI often necessitates human collection and processing of data generated in real-world industrial settings, such as robotic movements and manufacturing processes.
For instance, in industries like shipbuilding, where processes are often unstructured, existing data may be insufficient, necessitating the creation of new datasets on-site, which can be time-consuming and costly. Consequently, the data construction process, which requires a large workforce, has made labor costs a critical factor.
An industry insider noted, "As the Physical AI market expands, the role of human labor is becoming increasingly important. For example, the need for labeling in autonomous driving and robotics is driving companies to seek out countries with lower labor costs."
Developing countries like Vietnam are securing substantial contracts for data construction from global AI firms, fostering the growth of related industries. According to market research firm Statista, Vietnam's IT outsourcing revenue is projected to reach $1.28 billion (approximately 1.97 trillion won) by 2028, growing at an average annual rate of 17%. Notably, demand for data annotation services is increasing by 15% each year.
However, the growing dependence of the AI industry on low-wage labor raises ethical concerns. As Physical AI requires a higher proportion of human involvement in data collection and processing, it may lead to an even greater reliance on labor compared to traditional generative AI.
Choi Byung-ho, a research professor at Korea University's Human-Inspired AI Research Institute, stated, "The AI industry is increasingly reliant on the labor force of low-wage countries, leading to ongoing criticism regarding labor rights and ethical issues."
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.