The South Korean government will join a cyber security consultative body run by OpenAI. Along with Japan, it is the first case among Asian countries.

The Ministry of Science and ICT said on May 27 that it held a meeting the previous day with OpenAI to explore cooperation on responding to artificial intelligence (AI) security threats and securing AI safety and trust, and formalised its participation in the Government Trusted Access for Cyber (GTAC) program.

Officials from both sides attended, including Ryu Je-myung (류제명), second vice minister of the ministry, and OpenAI Chief Strategy Officer Jason Kwon (Jason Kwon).

GTAC is a cyber security consultative body OpenAI runs for governments and public institutions. Participants gain access to OpenAI's latest high-performance AI models. OpenAI said South Korea, along with Japan, is the first country in Asia to join GTAC. Working-level tasks will be handled by the Korea Internet and Security Agency (KISA).

The two sides agreed to continue discussions on ways to use AI models in cyber security, starting with participation in GTAC.

Cooperation on AI safety was also placed on the agenda. The ministry requested the establishment of a cooperative relationship between the AI Safety Institute and OpenAI to respond to risks arising from improved AI performance and expanded use. It said this was to pursue practical cooperation, including AI safety evaluations and joint research. OpenAI agreed to actively review it.

Ryu said, "This cooperation with OpenAI has laid the groundwork to prepare in advance for AI security threats," adding, "We will continue to enhance domestic AI security capabilities through active cooperation and working-level discussions with global AI companies."