South Korea joins OpenAI's trusted cyber access program

By Kim Dong-young Posted : May 27, 2026, 10:43 Updated : May 27, 2026, 10:43
OpenAI's logo shown on a mobile phone/ AFP-Yonhap
 
SEOUL, May 27 (AJP) - South Korea has become one of the first two countries in Asia, together with Japan, to join OpenAI's Trusted Access for Cyber program, securing government-level access to the U.S. firm's most advanced artificial intelligence models for cybersecurity research and defense.

The Ministry of Science and ICT announced Wednesday that Vice Minister Ryu Je-myung met OpenAI Chief Strategy Officer Jason Kwon in Seoul on Tuesday to formalize Seoul's participation in the program, which grants vetted governments and institutions-controlled access to OpenAI's frontier models for defensive cyber work.

The Korea Internet & Security Agency will serve as the implementing body, the ministry said, adding that the two sides agreed to continue talks on broader applications of AI in cybersecurity. The arrangement is expected to give Seoul a window into how rapidly advancing AI systems could be weaponized, and how to blunt that threat.

"Through this cooperation with OpenAI, Korea has laid the groundwork to get ahead of AI-driven cyber threats," Ryu said, pledging to deepen engagement with global AI companies to sharpen the country's defensive capabilities.

The two sides also discussed building a working framework between Seoul's AI Safety Institute and OpenAI for joint safety evaluations and research on increasingly capable models. OpenAI said it would actively review the proposal.

Separately, the Korean government is exploring participation in Project Glasswing, a rival coalition led by Anthropic that pairs its unreleased Claude Mythos Preview model with about a dozen launch partners — including Amazon, Apple, Google, Microsoft and Nvidia — and some 40 additional organizations to hunt down software vulnerabilities. No Korean entity has yet joined the initiative.

Copyright ⓒ Aju Press All rights reserved.