South Korea secures access to Anthropic’s Mythos AI model, Science Ministry says

South Korea has secured access to Anthropic's advanced cybersecurity AI model, Mythos, signaling a significant step in the nation's efforts to bolster its digital defenses. The Korea Internet & Security Agency (KISA) achieved this access by participating in Anthropic's Project Glasswing, an initiative designed to leverage cutting-edge AI for identifying and mitigating cybersecurity threats.

The Ministry of Science and ICT confirmed KISA's involvement in the project, emphasizing continuous collaboration with Anthropic. Project Glasswing aims to utilize frontier AI models to proactively detect and address vulnerabilities within digital systems. This collaboration positions South Korea at the forefront of adopting innovative AI solutions for national security.

This development follows a Financial Times report indicating Anthropic's expansion of Mythos access to approximately 150 organizations across more than 15 countries. Prominent South Korean corporations, including Samsung Electronics, SK Hynix, and SK Telecom, are reportedly part of this expanded network. While Samsung Electronics and SK declined to comment, the Ministry's confirmation solidifies South Korea's strategic engagement with this powerful AI tool.

South Korea plans to sustain its drive to enhance cybersecurity capabilities. This includes exploring and integrating various frontier AI models and strengthening domestic AI-driven information security technologies. The access to Mythos is a key component of this ongoing strategy, aiming to build a more resilient and secure digital infrastructure.