As artificial intelligence (AI) emerges as a strategic industry that can determine the growth of countries and corporations, major governments are actively working to attract AI talent. Judging that it is difficult to catch up with the AI powers of the United States and China within the "golden time" through domestic talent cultivation alone, major countries are exerting full efforts to secure "AI brains" by introducing systems to bring in AI experts from around the world.
Experts advise that Korea should minimize the outflow of competent domestic AI talent while providing support for overseas AI experts to work in Korea.
According to the industry on the 1st, AI talent capable of building and enhancing advanced AI models and planning AI services has emerged as a national strategic asset as important as AI infrastructure or data. However, the supply of AI specialists with such capabilities is grossly insufficient compared to demand, leading both corporations and the government to reach out to talented AI professionals regardless of nationality.
According to a recent report titled "Global AI Talent Flow" published by the Boston Consulting Group (BCG), the United States, which is leading the AI era, has seen a net influx of more than 32,000 foreign talents over the past three years. This accounts for about 7% of the total AI workforce in the U.S. (approximately 488,000 people). Many of these individuals work for U.S. tech corporations, making up about 40% of all AI-related positions.
It is known that among "S-class" talents recruited by big tech corporations with salaries ranging from hundreds of millions to billions, more are overseas nationals than locals. In the case of Meta, the parent company of Facebook, 28% of its entire AI workforce has foreign nationality. For AI companies xAI, led by Elon Musk, and Google, each stand at 27%, Amazon 25%, while OpenAI and Microsoft (MS) each stand at 24% and Anthropic at 17%.
Of the 44 employees at the Meta Superintelligence Labs (MSL) focused on the study of "superintelligence" that exceeds human intelligence, it has been revealed that about half, or 21, are Chinese AI researchers who graduated from prestigious universities in China such as Tsinghua University, Peking University, and Zhejiang University. In fact, 55% of those with Ph.D. degrees in computer science and mathematics in the United States are foreign nationals, with about 75% of them being from China or India.
So-called "AI middle-power countries," including the European Union (EU), the United Kingdom, Australia, and Japan, have begun actively working on building AI infrastructure and attracting AI talent to narrow the gap with the U.S. and China, which lead the global AI market.
In February, the EU allocated approximately $585 million (about 810 billion won) to attract foreign AI researchers over the next two years. France plans to invest over $100 million (about 140 billion won) to attract researchers working in the U.S. The U.K. is preparing a researcher attraction program worth $70 million (about 975 billion won), while the Australian Academy of Science has introduced a migration support package to attract the world's top scientists and engineers. Various financial and institutional benefits, including support for education and research expenses and job creation, are included.
BCG noted that "the strategies pursued by each country are attempts to newly design 'a landing place for AI researchers,'" adding that "AI talent is not just a hiring target, but a core asset that determines the innovation capabilities of nations and corporations."
Korea, which is facing challenges due to the outflow of competent AI talent, is also urged to develop similar policies. It was investigated that Korea is the fourth largest country among the 38 member states of the Organisation for Economic Co-operation and Development (OECD) in terms of net outflow of AI talent. According to the Paulson Institute at the University of Chicago, about 40% of AI master's and doctoral graduates left for overseas as of 2022. The phenomenon of medical school enrollment has intensified in recent years, leading to a shortage of people in the engineering and science fields.
Experts point out that the conditions for AI talent to settle in Korea include a robust AI infrastructure, such as AI data centers and graphics processing units (GPUs), which are essential for AI operations, as well as a suitable research environment, excellent colleagues to collaborate with, and appropriate treatment. Korea is analyzed to be lacking in these conditions, causing talent to leave for better opportunities abroad.
As the domestic AI ecosystem is not yet fully established, there are opinions that drastic support measures are needed to retain departing domestic talent and attract global AI talent. The industry claims that a "fast track" system should be introduced to quickly issue visas to excellent overseas AI talent and that research institutes should be established where AI researchers can focus on their work peacefully. They emphasize that given the difficulty of attracting talent with competitive salary offers similar to those of U.S. big tech companies, there should be a focus on creating a stable and autonomous research environment, rather than solely concentrating on attracting S-class talent, aiming to cultivate A-class teams equipped with the capabilities to apply AI in industrial settings.
Jang Jin-seok, head of AI & Digital at BCG Korea, remarked that "What matters now for Korea, an AI middle power, is execution capabilities—the ability to implement AI in real industrial settings." He added, "The government should establish policy frameworks to minimize talent outflow and attract global experts, while companies must adopt systematic strategies to cultivate entire A-level teams rather than relying solely on 'superstar' individuals."
The day before, the government launched the "Generative AI Leading Talent Cultivation Project," a program designed to train advanced talents who can be deployed in the industrial field, in collaboration with private corporations such as LG AI Research Institute and NC AI. However, it has raised concerns that the budget of 18 billion won is small compared to advanced countries and that additional follow-up measures are necessary.
Moreover, experts emphasize that the government should accelerate the construction of an "AI highway" and create an attractive market environment where AI talent would want to stay. The government is actively pushing forward with large AI support projects such as the "Independent AI Foundation Model Project," the "National AI Computing Center Establishment Project," and the "Strengthening of GPU Resource Utilization."
Japan, which has been evaluated as a latecomer in AI, recently stated its intent to become the "most AI-friendly country in the world" by enacting the "Act on Promotion and Utilization of Artificial Intelligence Technology Research and Development" and is implementing a policy to support corporations in purchasing GPUs in the form of subsidies.
Choi Byung-ho, a professor at Korea University's AI Research Institute, noted that "Japan may not have many outstanding AI talents, but it has become easier to sell recruitment by emphasizing GPU acquisition." He added that "Korea should establish AI research institutes where capable researchers can gather and focus on building AI infrastructure."