President Lee Jae-myung's victory lays the groundwork for implementing his ambitious yet loosely defined technology-related pledges, which could have sweeping implications for Koreas future in artificial intelligence and semiconductors. Critics, however, warn that the nation should exercise greater fiscal discipline, given that domestic industries are falling behind amid intensifying global competition.
koreajoongangdaily.joins.com/, Jun. 04, 2025 –
Lee’s headline-grabbing promise to create a 100-trillion-won ($72.6 billion) AI fund stands out as one of the largest state-backed investments, although details on how the funds will be raised and allocated remains hazy. The fund aims to elevate Korea’s AI capabilities to the level of the United States and China, with plans to build infrastructure, nurture talent, support research and development (R&D) — including securing and organizing quality data — and bolster nationwide digital literacy. Lee has also pledged to empower a government body operating directly under the president to oversee AI initiatives.
During his inauguration speech on Wednesday, Lee again promised to commit "extensive investment and support" for the two sectors, so that the nation "will rise as a powerhouse leading the industries of tomorrow."
However, Lee’s promises follow rather than lead the global trend. Several countries have already rolled out billion-dollar national AI infrastructure initiatives, with many investing heavily in energy capacity to support AI growth. To narrow the gap, Lee plans to accelerate AI adoption across all economic sectors by rolling out large-scale public AI projects aimed at delivering services nationwide.
Lee's vow to supply 50,000 graphics processor units is seen as a modest start, with experts suggesting Korea will need far more to remain competitive.
“For community-driven progress, we need a large enough cluster — like 100,000 GPUs — to enable experimentation and scale, rather than just equipping a few elite teams,” said Professor Billy Choi of Korea University’s Human-inspired AI Research Lab.
Beyond hardware, high-quality data is crucial for training cutting-edge AI models. Preparing public data for private use also demands significant investment to clean, structure and produce synthetic datasets.
Equally essential is talent, which Korea’s AI sector continues to lag behind as top talent continues to exit the country. The problem was further aggravated by the former Yoon Suk Yeol administration’s 14.7 percent cut in the 2024 science R&D project — the first reduction in 33 years.
Lee’s ambitious promises, however, risk faltering under scrutiny, as many of his proposals remain broad and lacking in concrete policy detail.
“Technology is advancing so rapidly that golden opportunities keep arriving, but social consensus always lags,” Choi said. “The government needs to step up decisively and lead. What makes things even more complicated is that this sector is deeply intertwined with national security. A misstep here could undermine both our industries and our security.”
Lee also promised to revitalize the semiconductor sector, which has suffered from oversupply and rising competition from Chinese firms. One key promise is to fast-track a special semiconductor law providing subsidies and easing the 52-hour weekly work limit. While Lee has historically aligned with labor unions in opposing extended working hours, he now proposes a 10 percent tax cut for domestic chip production.
Lee Jong-hwan, a professor at Sangmyung University’s Department of Semiconductor and Display Engineering and a former Samsung Electronics engineer, argued that the 52-hour cap is a lower priority. Rather, he said, the government should strengthen the entire chip supply chain — including support for the small- and medium-sized enterprises, as well as materials, components and equipment companies — by learning from TSMC’s model.
“Comprehensive support — including subsidies, infrastructure investment and work force development — is essential,” he said. “Governments in the U.S., Taiwan, China and Japan are all stepping in to help their companies. This means going beyond memory chips to system semiconductors, including chip manufacturing and advanced computer chips, which have even bigger markets. If we focus only on memory, we’ll miss out on huge opportunities like AI chips and autonomous driving.”