Home > Business > Tech

AI takes center stage as candidates unveil plans ahead of Korea's presidential election

Published: 30 Apr. 2025, 15:43

Semiconductor chips are seen on a circuit board of a computer in this illustration picture taken Feb. 25, 2022. [REUTERS/YONHAP]

AI has emerged as the most important campaign issue ahead of the June 3 presidential election. With the United States and China locked in a global race for AI supremacy, the technology is now seen as a key driver of national competitiveness.

Major presidential candidates are pledging investments amounting to hundreds of trillions of won to propel Korea into the ranks of global AI powerhouses.

Lee Jae-myung, the Democratic Party’s (DP) presidential candidate, has centered his platform around building an “AI Basic Society.”

On April 14, in his first official campaign activity following his candidacy announcement, Lee visited Korean AI semiconductor firm FuriosaAI and said, “We will usher in an era of 100 trillion won [$70 billion] in AI investment.”

He also pledged to secure at least 50,000 graphics processing units (GPUs), a key component of AI infrastructure. In addition, Lee promised to develop a “Korean version of ChatGPT” that would be “freely accessible to all citizens.”

Democratic Party presidential candidate Lee Jae-myung gives a speech in western Seoul on April 30. [YONHAP]

Back in March, Lee had already sparked debate by proposing the creation of a “Korean Nvidia,” saying, “Let’s build a company like Nvidia where the private sector holds 70 percent and the remaining 30 percent is shared among the people so it doesn’t rely on taxes.”

Candidates in the People Power Party (PPP) primary have also rolled out their own AI pledges.

Han Dong-hoon promised to invest 150 trillion won in infrastructure and 50 trillion won in the AI ecosystem, for a total of 200 trillion won. He emphasized strategic investment in applied areas, such as medical AI, robotics and military technology, stating that Korea could build its own version of Palantir. The proposal was seen as a direct counter to Lee’s “K-Nvidia” plan.

Kim Moon-soo pledged to create a 100 trillion won fund to support AI unicorn startups and train 200,000 young AI professionals. Lee Jun-seok, candidate of the Reform Party, said he would push for fair-use guidelines regarding AI training data, addressing growing conflicts over personal data and copyright.

But industry observers are skeptical of these big-number pledges, citing a lack of concrete planning and realistic funding measures.

Although the government doubled this year’s AI budget in a supplementary bill, it still amounts to just 3.6 trillion won. Experts warn that even half of the 100 trillion won promise would be difficult to spend within a single presidential term without viable funding sources.

People Power Party (PPP) presidential primary candidates Kim Moon-soo, left, and Han Dong-hoon on April 25 in western Seoul [JOINT PRESS CORPS]

“An investment of over 100 trillion won cannot be achieved through government research and development alone. Private sector and even global funds must be mobilized,” said Moon Jeong-hoon, professor of industrial engineering at Yonsei University.

He pointed out that the United States’ $700 billion Stargate AI infrastructure project includes not only private investments from OpenAI and SoftBank, but also oil money from the Middle East.

Korean experts also stress the need to attract foreign investment for projects like data center construction.

Others criticized the proposals for lacking strategic focus.

“The problem is that we don’t see any real industrial strategy." Kim Dong-hwan, CEO of AI startup 42Maru. "Do they want to build large foundational models centered on conglomerates? Or are they aiming to foster a startup-led AI transformation ecosystem to target global markets? Or do they plan to participate in global giants like OpenAI or Microsoft as strategic investors? A clear strategy is needed.”

Lee’s “Korean ChatGPT” proposal also called for caution.

A computer keyboard lit by a displayed cyber code is seen in this illustration picture taken on March 1, 2017. [REUTERS/YONHAP]

“While a government-led model may help generate AI demand, it could also slow down private sector innovation, just as state-led telecom projects lagged in the past," said Hong In-gi, professor of electronic engineering at Kyung Hee University.

Kwak Gyu-tae, professor of global cultural industries at Soonchunhyang University, added, “We should admit that we’re already behind in foundational AI technology and instead pursue a fast-follower strategy.”

The Digital Economy Alliance — comprising seven organizations including the Korea Internet Corporations Association and the Korea Venture Business Association — recently issued a policy proposal highlighting infrastructure and talent shortages as the private sector’s biggest pain points. The alliance called for concentrated support in these areas.

“It’s crucial to adopt a strategy of selection and concentration so that leading companies can drive the entire industry forward with the limited resources we have,” said Professor Kwak.

Translated from the JoongAng Ilbo using generative AI and edited by Korea JoongAng Daily staff.
BY YUN JUNG-MIN, JEONG YONG-HWAN [[email protected]]