Korea’s approach to the AI revolution

The interview was conducted by Ula Rutkowska (Senior Researcher, Apolitical) and edited by Christina Obolenskaya (MSc in International History, LSE and Communications Intern, Apolitical). Jaeho Yeom was named on Apolitical’s Government AI 100 2025.

Artificial intelligence is rapidly reshaping economies and societies, and governments around the world are racing to develop strategies that harness its potential while addressing its risks. In Korea, AI is seen as a force for improving lives, streamlining work, and driving innovation — but also as a technology that could deepen inequalities if not carefully managed.

Jaeho Yeom, Vice-Chair of the Presidential Commission on AI of Korea, outlines Korea’s vision for AI adoption, from building national infrastructure to addressing privacy concerns and ensuring responsible governance. In this interview, he discusses how Korea is investing in AI research, strengthening public-private cooperation, and preparing for the social and economic shifts that AI will bring. He also reflects on the challenges ahead—such as security risks, energy demands, and changing workforce dynamics—and how Korea is positioning itself to navigate them.

Q: What are the key pillars of a successful national AI strategy, and how do you ensure it remains forward-looking?

The first one is AI culture. There’s a kind of fear about AI, and some people feel that AI technology will dominate human society. It is worrying, but it’s similar to how, in the 15th century, printing was invented. Then, everything changed, making it possible for books and printed materials to be widely dispersed. As a result, the Reformation, Renaissance, Scientific Revolution, and even Industrial Revolution followed.

All humanities, philosophy, and the social system changed very rapidly. AI is the same kind of thing. We will just use AI in various ways, and it will help reduce our labour time and avoid unnecessary repetitive chores.

The second is data and privacy. Korea collects a lot of hospital and medical data, but most hospitals cannot reveal the data because of individual privacy issues. Revealing it anonymously is possible. So, we will need to develop the technology to gather and reveal the information and data. The underlying reason is that if the data is revealed, we can analyse hospital and doctor behaviour, diagnoses, treatments, and how they make money. Without data, it's impossible to use AI to improve these services.

Q: What are the critical elements of a strong AI infrastructure at the national level in Korea?

At the national level, the Korean government will invest around two billion US dollars to form an AI computing centre. Many researchers in universities and startups want to use AI but don't have enough resources or money to do so. We will provide the facilities and infrastructure. If anyone wants to try some trial learning processes with their big data and get some ideas from AI, we will provide that.

The second thing is that Korea is very strong in AI semiconductors. We will support new semiconductor technology along with the private sector, like Samsung or SK Hynix. We are expecting around 50 billion US dollars in investment. Around 50 kilometres from Seoul, in a suburban area, SK Hynix and Samsung are building new semiconductor factories, which will probably be completed by 2027 or 2030. It is a kind of ecosystem. Another thing is that AI needs a lot of energy – Korea is very strong in nuclear power plants, and since we have some unused power plants, we need to figure out how to utilise those facilities.

Q: What do you see as some of the biggest obstacles and challenges that governments face in shaping effective AI policy?

If human beings cannot control AI, it can be problematic. I read an article from CES where Jensen Huang from NVIDIA mentioned that it is not the machine itself that is the issue. It is human beings who control and manipulate AI, causing harm to others. AI machines cannot do that on their own, even if they improve their intelligence. We need to educate AI machines on how to behave properly, just like we educate children on morals and behaviour.

Another challenge is national security, such as hackers and terrorist attacks. Protecting against these threats is crucial. Additionally, there is the issue of disparity between AI users and non-users, leading to inequality. Several years ago, I didn't agree with universal income, but with AI support, productivity will go up, and society will be richer. This could allow for universal income, ensuring a basic income level for everyone.

Social issues will arise, and humanities and social systems will become more important. For example, family concepts are changing rapidly. In Korea, 10 or 20 years ago, only blood ties defined family. Now, broken families and even pets are considered family members. Human concepts and philosophy will change rapidly. Thanks to AI technology, new democratic voting systems and governance systems will evolve, just like political systems changed after the invention of printing.

Q: What does effective public-private cooperation look like?

Korea has a long history of industrial policy, with government support similar to Japan. However, after democratisation, many industrial people feel that the government is a stumbling block with tight labour regulations and controls. Despite this, top businesses like Samsung and SK Hynix contribute significantly to Korea's GDP. If they fail, it would be a serious danger to the economy. Even the opposition party has mixed feelings, supporting the industry but aligning with labour parties. We need to have conversations and compromises.

Many people see AI as a future industry. However, there are conflicts, such as the opposition to nuclear power plants despite the need for electricity. Despite political turmoil, I am optimistic that the opposition party will support the industries.

Q: How will AI be applied in various fields, including education, in the coming years?

AI will be applied everywhere, including homes, workplaces, and education. In Korea, this year, we will start using digital textbooks for elementary, junior high, and high school students in English, mathematics, and computer science. Teachers won't teach at all; students will learn through computer-assisted, AI-assisted teaching. Everything will change in 10 years. The young generation, especially startups, are very innovative.

I met a colleague, a professor at my former university, who is starting a new startup. Thanks to AI, bank loans, which are very difficult for small and medium companies and startups, can be analysed. He analysed banking data and startup owners' behaviours over 10 years to react quickly using AI data. This makes it easier to get loans from banks, with support from banks and telephone companies.

When I returned from Stanford to Korea as a professor forty years ago, everyone was focused on information or IT technology. We were lagging behind in industrialisation, but we moved ahead in information. Korea has a very good infrastructure, and AI will spread quickly this year. Everyone understands that AI is a new civilisation.