Public research body · Science policy · Pangyo Global R&D Center (Global R&D Center), Seongnam/Pangyo area · Est. 2024
The Korea AI Safety Institute (K-AISI) is South Korea’s government-backed research organization for AI safety research, evaluation methods, and international cooperation. It was established under the Electronics and Telecommunications Research Institute (ETRI) following the AI Seoul Summit (May 2024) as a dedicated hub to address AI risks arising from technical limitations, misuse, and loss of control, and to help build scientific and data-grounded policy capabilities.
Per Korea’s Ministry of Science and ICT (MSIT), K-AISI’s role includes (1) defining AI risks and developing safety evaluation approaches and tools, (2) conducting policy research aligned with international AI safety norms, (3) promoting cooperation among industry, academia, and research institutions through an “AI Safety Partnership,” and (4) conducting R&D to strengthen AI security/robustness and address threats such as deepfakes. Operationally, K-AISI is positioned at Pangyo’s Global R&D Center to facilitate recruitment and collaboration with AI companies and related institutions. MSIT also states the institute is a member of the International Network of AI Safety Institutes and is intended to contribute to global AI safety governance science and practice.
No indexed openings right now. Check the careers page ↗.
Publications are not yet indexed for this organization. Read them on msit.go.kr ↗.
Aju Press reported scrutiny directed at the Korea AI Safety Institute for allegedly failing to report assessments, including when asked for specific results; the article attributes a response that institute results are disclosed on its website.
NAIRL partners with Korea AI Safety Institute for AI innovation and safety (LOI signing)NAIRL announced that its directors and the Korea AI Safety Institute director signed a letter of intent during a signing ceremony at K-AISI headquarters in Pangyo (Apr 20, 2026), framed as AI innovation and safety collaboration.
Korea AI Safety Institute launched as Korea’s AI research hub (launch ceremony)MSIT held a launch ceremony for the AI Safety Institute at the Pangyo Global R&D Center (Nov 27). MSIT framed the institute as a dedicated organization to systematically and professionally address AI risks and to foster collaboration across industry, academia, and research institutions, including global network participation.
Professor Kim Myuhng-joo appointed as first director of the AI Safety InstituteMSIT announced the appointment of Professor Kim Myuhng-joo as the inaugural Director of the AI Safety Institute, effective Nov 12, and described his background in AI ethics and reliability; MSIT also provided the term dates (3 years).
AI Safety Institute plan approved/announced: functions, staffing, and international cooperationMSIT’s English press release describing the ICT Strategy Committee meeting states that, following the AI Seoul Summit (May 2024), the AI Safety Institute would be established in November under ETRI with three departments and described its missions and functions (risk evaluation, policy research, cooperation, and R&D).