Project Glasswing, a security consortium formed around Mythos for cybersecurity vulnerability detection, grants access to participating organizations including Samsung Electronics, SK hynix, and SK Telecom, while export controls threaten Korea’s AI threat-response capabilities.

Korea must build domestic AI capabilities for defense, public sectors, and national security, and push policies to develop independent AI foundation models and localized AI ecosystems.

The broader policy trajectory focuses on the localization of AI ecosystems, investments in independent foundation models, and a strategic push to accelerate sovereign AI initiatives to narrow the gap with the U.S. and China.

The U.S. has blocked foreign access to Mythos 5 and Fable 5 for overseas users, foreign nationals in the U.S., and Anthropic employees, prompting service suspensions by Anthropic and fueling industry concerns about AI as a strategic good.

Korea is responding to U.S. export controls on Anthropic’s Mythos by preparing countermeasures and assessing implications for national security and competitiveness.

Industry voices warn that broader access restrictions to other major AI models could affect corporate and national security, underscoring the urgency of sovereign AI investments.

Experts note Mythos specializes in vulnerability analysis rather than general-purpose AI, intensifying competition to secure sovereign AI capabilities and boosting momentum for local development.

Korea’s Ministry of Science and ICT and industry players, including SK Telecom, are coordinating response plans and maintaining communication with Anthropic as they assess the situation.