Translation result.
[Digital Today Reporter Son Seul-gi] Domestic artificial intelligence (AI) models are increasingly being deployed across public administration, from government R&D budget reviews to support services for older adults living alone.
On May 26, the Ministry of Science and ICT said it is integrating homegrown AI models developed by companies such as Upstage, LG AI Research, and Naver into central and local government operations.
Upstage’s proprietary AI model is being used in national R&D budget reviews. The system analyzes and organizes research proposals and budget details to improve the rigor and efficiency of evaluations. It performs similarity and duplication checks, automates administrative workflows, drafts initial documents, and supports collaborative review processes.
Domestic AI models have also been adopted on the government-wide platform known as the Government-wide AI Common Platform. Built jointly by the Ministry of Science and ICT and the Ministry of the Interior and Safety, the platform enables central and local governments to share and jointly use a variety of AI models.
Under the ‘K-Moonshot’ initiative, the government and private AI firms will work to develop science-focused AI models for strategic technology areas such as biotechnology and semiconductors. The effort targets scientific challenges in fields including drug discovery, next-generation semiconductors, and nuclear fusion.
This year the government will host a nationwide AI competition that uses K-AI models on the platform. The contest will run in four divisions—general public; elementary, middle and high school students; university students and researchers; and digitally vulnerable groups—and officials are considering pathways to commercialize standout projects and link them to startups and employment opportunities.
In public safety, LG AI Research’s EXAONE model will power an ‘AI Safety Hotline.’ In partnership with the Ministry of the Interior and Safety, a pilot service is planned this year to apply the system to disaster prevention, facility risk detection, and anomaly analysis.
Local governments are expanding their adoption as well. The city of Paju has introduced LG AI Research’s model for civil service and administrative tasks. Busan developed an ‘AI Accounts Clerk’ based on Naver’s AI model and has rolled it out across its municipal administration.
In elder care, Naver’s AI-driven ‘Naver Care Call’ places regular check-in calls to older adults living alone to assess their health and emotional well-being. The service currently reaches about 50,000 people through roughly 160 institutions, including the Seoul and Busan city governments and the Gyeonggi Provincial Social Services Agency. A study found the program generated about 34 billion KRW (approximately $25.5 million) in social value in 2025, and international academic research reported reductions in depression and improvements in memory function among people with dementia.
Ok Sang-hoon, leader of Naver Cloud’s Care Call business strategy, said, “We hope to reduce care blind spots and act as a social safety net that keeps people closer to support,” adding, “We plan to expand into various public-service areas, including local care and medical coordination.”
Most Commented