Artificial or intelligent? Lee Jae-myung's 100 trillion won AI pledge under scrutiny

Home > Business > Industry

Artificial or intelligent? Lee Jae-myung's 100 trillion won AI pledge under scrutiny

Published: 23 Apr. 2025, 07:00 Updated: 23 Apr. 2025, 13:45

LEE JAE-LIM
[email protected]

Audio report: written by reporters, read by AI

Presidential candidate Lee Jae-myung, left, and June Paik, CEO of chip designer Furiosa AI, shake hands for a photo during Lee's visit to the startup's office in southern Seoul on April 14. [JOONGANG ILBO]

[NEWS ANALYSIS]

When Lee Jae-myung, the front-runner in the upcoming presidential race, pledged to establish a 100 trillion won ($70.3 billion) fund to propel the nation's AI industry forward, reactions from industry insiders and experts were mixed.

While many have long called for large-scale, state-backed investments or subsidies in the rapidly evolving technology to close the widening gap with countries like the United States and China, they also expressed skepticism, pointing to past instances where government-announced “funds” were predominantly backed by private-sector capital or suffered from poor oversight and a lack of transparency.

Lee pledged on April 14 to establish a hybrid fund comprising 70 percent private investment and 30 percent public capital, aimed at fostering what he coined K-Nvidias, or globally competitive Korean AI firms that could be on par with the U.S. powerhouse.

To strengthen the country’s AI infrastructure, he also vowed to secure 50,000 GPUs and unveiled an ambitious plan to develop a “Korea-tailored ChatGPT,” which would be freely available to the public to rapidly generate large-scale data.

Is a 100 trillion won AI fund feasible?

Historically, no government-led funds have approached the scale of 100 trillion won. The largest-ever public-private investment initiative to date was the 20 trillion won New Deal Fund launched in 2020 under the Moon Jae-in administration to drive the nation’s digital and green transformation.

While several New Deal funds were fully deployed thanks to strong initial interest, the outcomes varied widely. Private equity and venture capital vehicles delivered more tangible returns, whereas public-participation index funds struggled, dragged down by market volatility in areas such as gaming and biotech.

Despite the scale of investment floated by candidates today, economists say it is not an unachievable number if the government is ready to act as a catalyst to promote the industry forward.

“Technically, raising that amount is possible,” said Prof. Lee Jeong-hwan at Hanyang University’s Department of Financial Economics.

“Outright fiscal spending — say, contributing 30 trillion won directly from the budget — is unrealistic, but the government can play a role through matching funds or seed capital. What matters most is that public money deployed this way must aim for returns, even if modest — around 1 to 2 percent. In order for this guarantee to be possible, the sector that it is investing in needs to stimulate enough market demand, which is the case for AI.

“AI is an extremely capital-intensive sector,” he added. “Companies are urging the government to step in, not because they lack ambition, but because they can’t mobilize capital quickly enough on their own.”

Looking for more than rhetoric

Industry watchers are closely watching to see whether the pledges reflect a genuine commitment to AI or mere populist rhetoric, as attention now shifts toward concrete investment plans.

“I think it helps when the goals are set boldly,” said an industry source from a startup who requested anonymity. “A 100 trillion won fund is certainly quite a large figure, but in the context of AI, Korea’s current scale is relatively small compared to other nations. What really matters in the end is who presents a detailed and concrete road map. So far, what we’ve seen is just the broad strokes of an outline.”

Aggregate private AI investment for a decade from 2013 was exceedingly led by the United States, totaling $335.24 billion, followed by China and Britain with $103.65 billion and $22.25 billion, respectively. Korea sat in ninth at $7.25 billion, according to The Stanford AI Index 2025 report published by the Stanford Institute for Human-Centered Artificial Intelligence.

Korea was 11th for the number of newly funded AI companies between 2013 and 2023 with 189, whereas the United States led the chart with 5,509, with China a distant second at 1,446.

Presidential candidate Ahn Cheol-soo of the People Power Party speaks at a press conference held in Daegu on April 21. [NEWS1]

In the political sphere, lawmakers with science and engineering backgrounds, such as People Power Party Rep. Ahn Cheol-soo, founder of cybersecurity firm AhnLab, and Reform Party Rep. Lee Jun-seok, who has a background in software development, have been at the forefront of criticism of AI pledges that emphasize large sums without concrete strategies.

The investment size alone, however, is meaningless without the establishment of clear industry guidance, a value system and implementation strategy, according to Prof. Billy Choi of Korea University’s Human-inspired AI Research Lab.

“Before debating budgets or regulations, the country must define whether it aims to replicate existing technologies or lead original innovation,” Choi said. “Korea, as of now, risks simply retracing China’s path in perhaps excelling in AI engineering but not innovating.

“In advancing AI infrastructure, it’s about more than just attaining some tens of thousands of GPUs,” he added. “That is a given, and two things are essential to boost innovation: a powerful, homegrown AI model to retain data sovereignty and to help local industries build AI-powered products, and quality data to boost the model. No foundational model functions without it. Korea has rich public data, but much remains inaccessible or poorly processed. It must be opened and made usable.”

Another industry insider from a tech company, speaking on condition of anonymity, told the Korea JoongAng Daily that the real impact of the pledges will hinge on how business-friendly the measures are.

"The pledges need to be translated into concrete, actionable plans that ease the burden on companies, especially in securing computing resources like GPUs, power and data centers," the source said.

BY LEE JAE-LIM [[email protected]]