GovMesh Digest: South Korea’s quest to move from Digital Platform Government to AI Government

South Korea's story is one of those featured in the GovMesh Digest special report. You can find the individual stories on the other participating governments at GovMesh 1.0 here. 

The Korean wave signals South Korea’s knack for global influence, with K-pop culture appeal having taken the world by storm.  

Now, the same innovative spirit is driving a new ambition: The K-AI Based government of Korea. 

At GovMesh, the National Information Society Agency (NIA)’s Executive Principal, Yoo Jae-sin, shared that NIA believes the next phase of digital government is an artificial intelligence (AI) government. 

GovMesh is a by-invite-only event for emerging digital governments in the region, and its inaugural edition happened in March 25 in Singapore. 

Yoo said that AI will become an essential innovation tool in the public sector.  

In his presentation, he explained the government’s evolution from a basic E-Government, which simply digitise public services, to a Digital Platform Government, a national strategy focused on collaborative service delivery and transforming government operations. 

Building on its impressive digital government journey (ranking first in OECD Digital Government Index for the second time), NIA is working on a generative AI (GenAI) common foundation for the government. 

The shared AI infrastructure would enable all public agencies to conveniently and safely tap into the technology to improve public services. 

Prioritising private sector collaboration 

To build this foundation, Yoo highlighted that private sector collaboration would be key.  

Specifically in its Guidelines for the Adoption and Utilisation of Large-Scale AI in the Public Sector published in April 2024, the key principles stated the need to prioritise the adoption of private sector cloud and AI solutions and promote public-private collaboration.  

The emphasis on public-private collaboration was evident in NIA’s previous work to make open government data more accessible and building the capacities of private companies to leverage open data for AI innovation. 

According to Yoo, the foundation operates on cloud infrastructure. “Private cloud service providers will operate within the public-private partnership (PPP) zone infrastructure,” he explained. 

In this model, the government provides the physical environment, and private cloud service providers (CSPs) deliver their services within it.  

To manage resources efficiently, the government employs a usage-based subscription fee for accessing Graphics Processing Unit (GPU) power. 

 
On top of a public-private cloud infrastructure, the government also taps on large language models (LLMs) and AI platforms offered by private South Korean companies. 

According to Yoo, government departments can only use South Korean LLMs and AI platforms as per National Intelligence Service (NIS)’ guidelines. 

These platforms operate on a usage-based subscription model. 

Current AI POCs in the government 

Yoo presented the emerging use cases of GenAI in the government, acknowledging that many are still in the proof-of-concept (POC) phase to address challenges related to data security and full implementation. 

The use cases could be categorised into improving operational efficiency and public service delivery.  

On improving operational efficiency, the Public Procurement Service is a GenAI tool that helps public officers draft requests for proposal (RFP) documents. 

Another example is the Patent Examination Support Service that is used in the Korean Intellectual Property Office. 

To enhance the efficiency and accuracy of the legitimisation process, the GenAI tool includes features to search for regulations and similar patents, as well as to summarise opinion statements.  

On improving public service delivery, Yoo shared about an award-winning POC from a competition organised by the Ministry of Interior and Safety to develop AI use cases in the public sector. 

Public transport companies are also able to use a ChatGPT-powered chatbot to train new employees and improve customer service. 

Another use case is a mobile app that taps on AI to deliver tailored support for young farmers in rural communities. This app includes a chatbot service, personalised notification service, educational information, and more.  

Finally, public officers at the Ministry of Employment and Labour could also use a GenAI tool to support labour law consultation and inspections.