[AI Talent Lost to Big Tech ⑤] Easing Visa Rules and Startup Acquisitions Fuel Global Talent Race

As countries around the world compete fiercely to recruit top artificial intelligence (AI) talent, there are growing calls for bold institutional reforms in South Korea to prevent the outflow of domestic AI professionals and to actively attract overseas talent.

원본보기 아이콘

According to the IT industry on July 18, Japan, which has become a net importer of AI talent since 2020, is implementing the "J-Skip" (Japan Special Highly Skilled Professional) visa system. This system provides preferential treatment to outstanding foreign professionals who meet certain educational, career, and income requirements. Benefits include spousal employment, permission to hire domestic helpers, and relaxed requirements for permanent residency, all aimed at attracting highly skilled workers. According to the Organisation for Economic Co-operation and Development (OECD), in 2023, South Korea experienced a net outflow of 0.3 AI technology holders per 10,000 people, while Japan saw a net inflow of 0.54 during the same period.

The Ministry of Justice launched the "Top-Tier Visa" system in April to attract outstanding talent in advanced fields. However, there are criticisms that the requirements are too stringent, as applicants must have obtained a master's or doctoral degree from one of the world's top 100 universities. Seo Yongseok, a professor at KAIST, told Asia Economy in a phone interview on the same day, "It is necessary to analyze the status of outstanding talent by field in each country and match it with domestic demand." He added, "We should expand special admissions for gifted and science high schools for foreign talent to support long-term residency, and also relax the criteria for dual citizenship to attract skilled technicians."

In addition to support measures aimed at preventing talent outflow, some point out that South Korea should benchmark the management strategies of big tech companies, which are desperately securing AI talent for their survival. One such strategy is the "acq-hire" approach, which involves acquiring promising startups outright or recruiting their key personnel.

For example, Google recently hired key researchers from the AI coding startup WindSurf, including CEO Varun Mohan and co-founder Douglas Chen. Instead of acquiring equity or controlling management, Google hired a small group of employees into Google DeepMind and secured some of WindSurf's technology through licensing. Meta, which has been actively absorbing AI talent, recently acquired a 49% stake in the AI startup ScaleAI and brought in its founder, Alexander Wang. At Amazon, David Luan, founder of the AI startup Adept acquired last year, now leads the company's Artificial General Intelligence (AGI) research lab.

원본보기 아이콘

Cho Sungbae, professor of computer science at Yonsei University, said, "The reason big tech companies sweep up talent is not just about money. They also offer opportunities to participate in challenging projects at leading companies equipped with cutting-edge AI research environments." Professor Cho added, "While domestic AI transformation (AX) companies recognize the need to adopt AI technologies, there are still few areas within their businesses where they can actually absorb and effectively utilize this talent. We should not approach the development of high-level talent by focusing only on isolated aspects without building a comprehensive AI ecosystem."

Experts also emphasize the need to bridge the gap between the capabilities that companies seek in AI talent and the actual skills that AI professionals possess. As the pace of AI technology development accelerates, new trends continue to emerge, but the technical level of domestic talent is said to be lagging behind. An AI service developer, who requested anonymity, said, "While the latest AI trends are rapidly shifting toward agentic AI and physical AI, domestic AI professionals are still focused almost exclusively on developing large language models (LLMs), to the point of calling it an 'LLM boom.' In the field, there is a growing need for talent capable of developing specialized technologies such as MLOps or context engineering, as well as individuals with diverse academic backgrounds beyond computer science."

© The Asia Business Daily(www.asiae.co.kr). All rights reserved.