Shinsegae, Reflection AI to Build Korea’s 250 MW Sovereign AI Data Center
In San Francisco on the 16th, Shinsegae Group of South Korea and Reflection AI announced a strategic partnership via a memorandum of understanding to establish a Korean sovereign AI factory. The signing took place at the National AI Center in San Francisco.
Under the agreement, the two companies plan to build a large-scale AI data center in Korea with a capacity of about 250 megawatts. The project is described as a flagship effort within Korea to advance sovereign AI capabilities and local data processing infrastructure.
Reflection AI, founded in 2024 by Misha Raskin (CEO) and Ioannis Antonoglou (CTO), traces its roots to developers from Google DeepMind and AlphaGo, positioning the startup at the intersection of advanced AI research and practical deployments.
U.S. participants at the ceremony included Howard Lutnick, identified as the U.S. Secretary of Commerce in the report. He emphasized a division of labor in the proposed collaboration: the United States would provide AI chips and models as well as full-stack engineering, while Korea would handle design, permitting, and financing. The officials framed the agreement as illustrating a new model for sovereign AI development.
Chairman Chung Yong-jin of Shinsegae Group described the partnership as just the beginning, expressing hope that the project would serve as a blueprint for nations that believe AI sovereignty and openness can coexist. Reflection AI’s CEO Laskin likewise said the initiative could become a blueprint enabling allied countries to secure AI sovereignty.
Shinsegae Group aims to leverage its retail expertise alongside AI capabilities to create differentiated AI agent commerce and to develop a full-stack AI platform for its retail businesses. The company says the effort could drive delivery innovations and help usher in an era of what it terms Emart 2.0, the next phase of its supermarket business.
Beyond Korea, the collaboration matters for U.S. readers because it touches on the global push to expand AI infrastructure, sovereignty over AI design and data, and cross-border partnerships that influence supply chains, technology policy, and regional competitiveness. The deal aligns with U.S. interests in fostering trusted AI ecosystems with allied economies while managing export controls, chip and model provisioning, and the development of secure, scalable AI capabilities.