"The key task for becoming a Top 3 in Artificial Intelligence is to foster a 'Korean OpenAI or Anthropic'" [Interview]

"The concrete outcome of the global 'Top 3 in Artificial Intelligence' goal is to foster K-AI companies that Koreans can proudly present to the world, just like OpenAI or Anthropic."
In an interview with The Financial News on the 9th, Lim Moon-young, vice chair of the National AI Strategy Committee, pointed to nurturing K-AI companies as the tangible achievement of the AI G3 Policy Objective. "Growing an industry ultimately means growing companies," he noted, adding, "The concrete picture of becoming a Top 3 in AI is that K-AI companies emerge that are even larger than Korea’s flagship exporters in semiconductors, petrochemicals, steel, shipbuilding and automobiles, and that people can clearly feel that Korea’s leading industry has shifted to AI."
Lim continued, "Just as the enterprise value of AI startup OpenAI, which has not even gone public yet, is estimated at 840 billion dollars (about 1,254.288 trillion won), and Anthropic at 380 billion dollars (about 567.682 trillion won), Korea has more than enough potential to produce K-AI companies that can be recognized in the global market," stressing, "Policy will build a structure that can support this." He added, "Unicorn startups such as Rebellions, FuriosaAI and Upstage are already in the market," and emphasized, "At a moment when timing is this perfect, we will align corporate goals with government support so they can produce results together."
It has been six months since the presidential National AI Strategy Committee was launched last September and Lim was appointed as its full-time vice chair. In January, with the enforcement of the Framework Act on the Development of Artificial Intelligence (AI Basic Act), the committee was converted into a statutory body, laying the groundwork for policy coordination and implementation. Over the past six months, based on discussions in each subcommittee, the National AI Strategy Committee finalized the "Korea AI Action Plan" last month. Consisting of 99 implementation tasks and 326 policy recommendations, it will serve as a roadmap for whole-of-government AI policy over the next three years.
■ "The AI Action Plan is not a declaration but an execution strategy"
Lim stressed, "AI policy is no longer at the stage of declarations or strategy announcements," and said, "The most important role of the National AI Strategy Committee is to turn the national strategy into actual execution, and the 'Korea AI Action Plan' is precisely an execution-focused strategy."
In particular, the Korea AI Action Plan focuses on realigning fragmented policies. "We conducted a full review of each ministry’s projects, streamlined overlapping tasks, and assigned priorities to core areas that the nation must concentrate on," Lim added.
To ensure the Korea AI Action Plan does not remain a mere declaration, an execution management framework was designed in parallel. Lim explained, "For each task, we will clearly designate who is responsible and set a deadline, and we will regularly check on implementation," adding, "The outcomes of the Korea AI Action Plan will be reported to the president and the public at the end of this year through the State Council of South Korea and the National AI Strategy Committee, and will also be reflected in the government performance evaluation conducted by the Office for Government Policy Coordination." He emphasized, "We are building a structure in which AI policy is managed based on outcomes," noting that the results of the Korea AI Action Plan will affect not only ministry-by-ministry evaluations but also each ministry’s budget allocations.
■ Building a one-stop site for all civil petition services
Lim revealed that the government is preparing an integrated AI civil service portal where all government civil petition services can be handled on a single site. "We will create a signature AI Government service that allows people to access all public services—such as welfare, civil petitions, military service and criminal matters—within one site," he said, explaining, "Because this project requires significant time and cost, the Public AI Transformation Subcommittee of the National AI Strategy Committee is doing extensive groundwork."
The National AI Strategy Committee has proposed "non-application-based services" as a core principle for AI-based public services. Under this model, government AI proactively provides welfare and civil services even if citizens do not apply. Lim pointed out, "In areas where the government needs information, such as taxes and military service, it already uses citizens’ data to impose obligations, but in welfare, people still have to apply one by one under the current system," and said, "We plan to use AI so that welfare benefits can be provided even if people do not apply." He added, "To realize non-application-based welfare using AI, the government must secure citizens’ personal data, so social consensus has to come first, and we expect services will be offered on an OPT OUT basis," explaining, "On that premise, we are reviewing ways to apply non-application-based services first in government services related to welfare and vulnerable groups."
■ Industrial convergence means policy silos must go
Lim criticized the tendency of ministries to build walls instead of cooperating, even as industries rapidly converge around AI. "AI is quickly integrating and transforming industries, and policy goals are also changing frequently, so ministries have no choice but to work together," he said. "The National AI Strategy Committee serves as the central control tower that sets the standards for the national AI strategy and aligns policies, so that each ministry’s policies do not overlap or move in conflicting directions."
A prime example is innovation in AI Government Infrastructure Governance. The authority over AI Government Infrastructure, which will underpin the Lee Jae-myung administration’s plan to convert public-facing services to AI, is a critical issue that cuts across every aspect of daily life. However, a fire last year at the National Information Resources Service Daejeon Center eroded public trust in the government’s infrastructure management. In response, the National AI Strategy Committee proposed measures including: raising safety standards for data centers in government and the public sector to at least the level of the private sector; transferring sensitive data to private cloud services; and establishing a dedicated cross-ministerial body under the Deputy Prime Minister for Science and Technology to oversee AI Government Infrastructure. Lim explained, "We designated the Deputy Prime Minister for Science and Technology as the person with ultimate responsibility for AI Government Infrastructure, and set out specific cooperative and implementation tasks for each project." He added, "The 'Manufacturing 2030 Plan' using AI will soon be announced, and for the AI Transformation (AX) of the manufacturing industry, including Physical AI, we will present detailed joint projects involving three ministries: the Ministry of Science and ICT, the Ministry of Trade, Industry and Resources (MOTIR), and the Ministry of SMEs and Startups."
■ "The pre-use copyright policy applies only where there is no existing market"
Regarding the controversy over copyright for AI training data under the "use first, compensate later" principle, Lim clarified, "It is a misunderstanding to think that pre-use and post-compensation will apply to all works," and stated, "For works where a transaction market exists, we will support a direction in which reasonable contracts and compensation are provided." He went on, "Where the copyright holder is unclear or no transaction market has formed, we are approaching the issue by applying different standards, such as allowing rights holders to exercise a right to refuse," and stressed, "Above all, social consensus is crucial when it comes to copyright issues for AI training data."
Lim concluded by defining the government’s role in the AI era: "The government will be the firmest pillar so that citizens and companies can confidently take on challenges in AI."
cafe9@fnnews.com Lee Gu-soon Reporter