Anthropic expands Project Glasswing to 150 organizations including Korean firms and government

Anthropic logo [REUTERS/YONHAP] AI company Anthropic has significantly broadened access to Claude Mythos Preview, its AI model capable of detecting security vulnerabilities at an expert level, expanding both the countries and organizations that can use it. Anthropic announced Tuesday that it is adding approximately 150 new organizations across more than 15 countries to Project Glasswing, its cybersecurity cooperation initiative built around Claude Mythos Preview. The project was launched after Anthropic released Mythos in April amid concerns that hackers could misuse the model. The newly added members span sectors not previously included in the program, such as power, water, health care, telecommunications and hardware. Related Article Samsung, SK hynix back Anthropic in $65B round, paving way for chip deal Anthropic taps 30-year industry vet Choi Ki-young as chief of operations in Korea The butterfly effect of the Anthropic contract termination "What each partner has in common is that a successful attack on their codebase could be catastrophic," it said. "For most partners, we estimate that a major attack could affect more than 100 million people, with important ramifications for both global and national security." Among the new joiners to the program are reportedly Samsung Electronics, SK hynix, and SK Telecom — three Korean firms that have recently invested in Anthropic. Samsung Electronics and SK hynix had participated in Anthropic's Series H funding round as "strategic infrastructure partners," making their inclusion in Project Glasswing a natural extension of that relationship. The Ministry of Science and ICT has also been included through the Korea Internet & Security Agency, the country's primary cyber incident response body. With both the government and major corporations joining Glasswing simultaneously, AI-powered cybersecurity vulnerability detection is expected to become more active in Korea. As concerns grow over AI-enabled cybersecurity threats, cooperation between AI companies and government and private sector organizations is on the rise. Last month, the Ministry of Science and ICT held a high-level meeting with OpenAI on responding to AI security threats and formally joined OpenAI's Government Trusted Access for Cyber program, which gives member organizations access to OpenAI's latest high-performance AI models, including GPT-5.5-Cyber. This article was originally written in Korean and translated by a bilingual reporter with the help of generative AI tools. It was then edited by a native English-speaking editor. All AI-assisted translations are reviewed and refined by our newsroom. BY KIM MIN-JEONG [lee.jian@joongang.co.kr]