Governments worldwide face a daunting challenge: understanding and managing the risks of rapidly evolving AI systems. As the technology advances at breakneck speed, risks emerge across both technical and strategic dimensions, demanding unprecedented collaboration across diverse stakeholder groups to build effective safeguards.
IST’s AI Risk Reduction Initiative tackles this challenge head-on. By convening developers, deployers, national security professionals, and policymakers, the initiative maps both the opportunities and risks of frontier AI and designs corresponding mitigation strategies. A recent collaboration between IST and the Korea AI Safety Institute (AISI) exemplifies how international cooperation can advance our understanding of AI risks across multiple domains.
The Korea AISI will soon release its AI Risk Map identifying distinct risk factors to help national stakeholders anticipate and address the hazards of rapid AI deployment. By organizing risks according to their causes, stages, and actors, the map provides governments and industry with a framework for developing targeted prevention and oversight standards. This timely effort responds to mounting AI-related incidents and growing concerns over diminishing human control of autonomous systems.
The IST-Korea AISI collaboration began in July 2025 with focused discussions on key components of the AI Risk Map. When the Korean delegation visited Washington in November, both organizations exchanged insights on their parallel efforts and identified concrete opportunities for deeper collaboration. The delegation showcased Risk Map progress and proposed joint work on mapping AI risks through a national security lens—an area where both organizations bring complementary expertise.
“We’re keen to learn from our Korean counterparts on their approach to AI risk and policies that support AI security and safety,” said IST’s Chief Trust Officer Steve Kelly. “For IST, this represents an important opportunity to see our work operationalized within an allied government framework.”
“We believe that working with IST, which has deep expertise in AI security, brings meaningful value to building a more systematic and practical AI risk map. We hope the partnership between IST and Korea AISI will contribute to reducing AI risks in the future,” mentioned Korea AISI’s Assistant Director, Dr. Minnseok Choi.
With support from the Patrick J. McGovern Foundation, IST continues engaging stakeholders across the AI ecosystem—including policymakers in the U.S. and internationally—to advance strategic priorities in AI policy and governance. Building on this foundation, IST and the Korea AI Safety Institute plan to deepen their collaboration through dedicated work on AI and national security risk mapping.
