[AI Talent Development Red Light] "US AI Engineers Earn 7-Figure Salaries... We Also Need Radical Salary Support"

Only One-Tenth of U.S. Salaries: Korean Companies Struggle to Secure Talent
Government Salary Matching and Easing of 52-Hour Workweek Needed as Dual Support Measures

In the artificial intelligence (AI) industry, there are proposals for the government to consider matching salary support to recruit not only professors but also top-tier AI talent. It is known that big tech companies in the U.S. offer so-called 'Seven Figures' (annual salaries of $1 million or more) financial compensation to engineers experienced in AI projects.

On the 10th, Kim Seong-hoon, CEO of AI startup Upstage, which developed the open-source model 'Sola,' said, "There are many Korean researchers active in Silicon Valley, such as at OpenAI and Anthropic," adding, "We tried to recruit them, but the salary gap is about tenfold, reaching around 2 billion KRW, which is beyond our capacity." He continued, "If the government matches about half of that, around 1 billion KRW, AI models surpassing OpenAI could emerge in Korea as well."

To nurture Korea's artificial intelligence (AI) talent, radical incentives like 'carrots' are necessary, and the industry has come forward with more concrete measures.

This proposal comes after observing that China's AI achievements are bearing fruit through active talent recruitment and infrastructure investment. Tsinghua University established the computer science experimental class called 'Yao Class.' Professor Yao Qizhi, who created the Yao Class, held professorships at major U.S. universities such as MIT and Stanford, but returned to his homeland in 2004 after receiving an unprecedented offer from the Chinese government promising "anything you want."

Moon Song-cheon, emeritus professor at KAIST Business School, explained, "It is difficult for Korean universities to promise such treatment when recruiting professors," adding, "Potential AI talents go abroad to learn under excellent professors, and even if they become smart scholars through that education, it becomes difficult for them to return home, creating a vicious cycle."

There are also criticisms that rigid working hours are holding back AI talent. Some suggest that exempting AI researchers from the 52-hour workweek could partially prevent brain drain overseas. Lee Kyung-jeon, professor of Big Data Applications at Kyung Hee University, said, "Due to the nature of AI research, the process of training models must be continuously monitored," adding, "This inevitably affects researchers' competitiveness. Do you think the DeepSeek development team complied with the 52-hour workweek while developing?"

Ha Jung-woo, head of Naver Future AI Center, said, "To prevent talent outflow, AI researchers themselves must be allowed to set their own time to focus on research," pointing out, "When paper deadlines approach and the 52-hour limit ends causing computers to shut down, researchers have no choice but to work secretly at home. This situation fails to recognize their labor."